clark-gates-george

3081 results

Rethinking Information Retrieval in MongoDB with Voyage AI

The future of AI-powered search The role of the modern database is evolving. AI-powered applications require more than just fast, scalable, and durable data management: they need highly accurate data retrieval and intelligent ranking, which are enabled by the ability to extract meaning from large volumes of unstructured inputs like text, images, and video. Retrieval-augmented generation (RAG) is now the default for LLM-powered applications, making accuracy in AI-driven search and retrieval a critical priority for developers. Meanwhile, customers in industries like healthcare, legal, and finance need highly reliable answers to power the applications their users rely on. MongoDB Atlas Search already combines keyword and vector search through its hybrid capabilities. However, to truly meet developers’ needs and expectations, we are expanding our focus to integrating best-in-class embedding and reranking models into Atlas to ensure optimal performance and superior outcomes. These models enable search systems to understand meaning beyond exact words in text, and to recognize semantic similarities across images, video, and audio. Embedding models and rerankers empower customer support teams to quickly match queries with pertinent documents, assist legal professionals in surfacing key clauses within long contracts, and optimize RAG pipelines by retrieving contextually significant information that addresses users’ queries. MongoDB is actively building this future. In February, we announced the acquisition of Voyage AI , a pioneer in state-of-the-art embedding and reranking models. With Voyage’s leading models and Atlas Search, developers will get a unified, production-ready stack for semantic retrieval. Why embedding and reranking matter Embedding and reranking models are core components of modern information retrieval, providing the link between natural language and accurate results: Embedding models transform data into vector representations that capture meaning and context, enabling searches based on semantic similarity rather than just keyword matches. Reranking models improve search accuracy by scoring and ranking a smaller set (e.g., 1000) of documents based on their relevance to a query, ensuring the most meaningful results appear first. A typical system uses an embedding model to project documents into a vector space that encodes semantics. A nearest neighbor search provides a list of documents close to a given query. These results are processed with a reranking model that enables deeper, clause-by-clause comparison between the queries and the nearest neighbors. This combination can greatly improve retrieval accuracy. For example, the system processing a user query for “holiday cookie recipes without tree nuts” may first retrieve a set of holiday recipes with the nearest neighbor search. In reranking, the query would be fully compared to each retrieved document to ensure each recipe does not contain any nuts. Voyage AI’s embedding and reranking models Voyage offers a suite of embedding models that support both general-purpose use cases and domain-specific needs . General models like voyage-3 , voyage-3-large , and voyage-3-lite handle diverse text inputs. For specialized applications, Voyage provides models tailored to domains like code ( voyage-code-3 ), legal ( voyage-law-2 ), and finance ( voyage-finance-2 ), offering higher accuracy by capturing the context and semantics unique to each field. They also offer a multimodal model ( voyage-multimodal-3 ) capable of processing interleaved text and images. In addition, Voyage provides reranking models in standard and lite versions , each focused on optimizing relevance while keeping latency and computational load under control. Voyage’s embedding models are designed to optimize the two distinct workloads required for each application, and our inference platform is purpose-built to support both scenarios efficiently: Document embeddings are created for all documents in a database whenever they are added or updated, capturing the semantic meaning of the documents an application has access to. Typically generated in batch, they are optimized for scale and throughput. Query embeddings enable the system to effectively interpret the user's intent for relevant results. Produced for a user's search query at the moment it's made, they are optimized for low latency and high precision. Figure 1. Voyage AI's embedding workflow: Document and query processing in MongoDB. Voyage AI’s embedding and reranking models consistently outperform leading production-grade models across industry benchmarks. For example, the general-purpose voyage-3-large model shows up to 20% improved retrieval accuracy over widely adopted production models across 100 datasets spanning domains like law, finance, and code. Despite its performance, it requires 200x less storage when using binary quantized embeddings. Domain-specific models like voyage-code-2 also outperform general-purpose models by up to 15% on code tasks On the reranking side, rerank-lite-1 and rerank-1 deliver gains of up to 14% in precision and recall across over 80 multilingual and vertical-specific datasets. These improvements translate directly into better relevance, faster inference, and more efficient RAG pipelines at scale. MongoDB Atlas Search + Voyage AI models today MongoDB Atlas Vector Search enables powerful semantic retrieval with a wide range of embedding and reranking models. Developers can benefit from using Voyage models with Atlas Vector Search today, even before the deeper integration arrives. Figure 2. Example code for embedding and vector search with Voyage AI and MongoDB. “AI-powered search”, not “AI Search” Not all AI search experiences are created equal. As we begin integrating Voyage AI models directly into MongoDB Atlas, it’s worth sharing how we’re approaching this work. The best solutions today blend traditional information retrieval with modern AI techniques, improving relevance while keeping systems explainable and tunable. AI-powered search in MongoDB Atlas enhances traditional search techniques with modern AI models. Embeddings improve semantic understanding, and reranking models refine relevance. But unlike opaque AI stacks, this approach remains transparent, customizable, and efficient: More control: Developers can tune search logic and ranking strategies based on their domain. More flexibility: Models can be updated or swapped to improve on an industry-specific corpus of data. More efficiency: MongoDB handles both storage and retrieval, optimizing cost and performance at scale. With Voyage’s models integrated directly into Atlas workflows, developers gain powerful semantic capabilities without sacrificing clarity or maintainability. Building the MongoDB + Voyage AI “better together” story While MongoDB’s flexible query language unlocks powerful capabilities, Atlas Vector Search can require thoughtful setup, especially for advanced use cases. Users must select and fine-tune embedding models to fit specific use cases. Additionally, they must either rely on serverless model APIs or build and maintain infrastructure to host models themselves. Each insert of new data and search query requires independent API calls, adding operational overhead. As applications scale or when models need updating, managing these new data types in clusters introduces additional friction. Finally, integrating rerankers further complicates the workflow by requiring separate API calls and custom handling for reordering results. By natively bringing Voyage AI's industry-leading models to MongoDB Atlas, we will eliminate these burdens and introduce new capabilities that empower customers to deliver highly relevant query results with simplicity. MongoDB is actively integrating Voyage's embedding and reranking models into Atlas to deliver a truly native experience. These deep integrations will not only simplify the developer workflow but will also enhance accuracy, performance, and cost efficiency - all without the usual complexity of tuning disparate systems. And our ongoing commitment to partnering with innovative companies across AI and tech ensures that models from various providers remain supported within a collaborative ecosystem. However, adopting the native Voyage models allows developers to focus on building their applications while achieving the highest quality of information retrieval. Figure 3. Enhanced AI-powered retrieval with MongoDB and Voyage AI. As we work on these native integrations, we're actively exploring advanced capabilities to further enhance the Atlas platform. Our investigations focus on: Defining the optimal approach to multi-modal information retrieval, integrating diverse inputs like text and images for richer results. Developing instruction-tuned retrieval, which allows concise prompts to precisely guide model interpretations, ensuring searches align closely with user intent. For example, enabling a search for “shoes” to prioritize sneakers or dress shoes, depending on user behavior and preferences. Determining the best ways to integrate domain-specific models tailored to the unique needs and use cases of industries such as legal, finance, and healthcare to achieve superior retrieval accuracy. Making it easy to update and change models without impacting availability. Bringing additional AI capabilities into our expressive aggregation pipeline language Improving the ability to automatically assess model performance, with the potential to offer this capability to customers. Building the future of AI-powered search From RAG pipelines to AI-powered customer experiences, information retrieval is the backbone of real-world AI applications. Voyage’s models strengthen this foundation by surfacing better documents and improving final LLM outputs. We are building this future around four core principles, with accuracy at the forefront: Accurate: ensuring the precision of information retrieval is always our top priority, empowering applications to achieve production-grade quality and mass adoption. Seamless: built into existing developer workflows. Scalable: optimized for performance and cost. Composable: open, flexible, and deeply integrated. By embedding Voyage into Atlas, MongoDB offers the best of both worlds: industry-leading retrieval models inside a fully managed, developer-friendly platform. This unified platform allows models and data to work together seamlessly, empowering developers to build scalable, high-performance AI applications with precision at their core. Join our MongoDB Community to learn about upcoming events, hear stories from MongoDB users, and connect with community members from around the world.

April 24, 2025

Reimagining Legacy Systems with AI: Why We're Building the Future

This article was adapted from an interview with Galileo’s Chain of Thought podcast. Watch the full episode on YouTube. At MongoDB, we talk a lot about what it means to be at an inflection point—a moment where you can either maintain the status quo or redefine what's possible. For my team and for software engineers around the world, that inflection point is here. Now. Large language models (LLMs), agents, and now model context protocol (MCP) are fundamentally changing not just how we work, but what is possible with technology. So we must adapt; we have to build differently, and be faster. At MongoDB, we're creating something unique. Something smarter. We’re embracing and exploring everything that AI can do for developers. And we’re inviting the next generation of engineers to join us in shaping it. The hidden cost of legacy systems If you've spent any time in enterprise engineering, you know the challenge many organizations face: decades-old systems that are critical to a company’s operations but are held together with duct tape and wishful thinking. The developers who built them have long since moved on. The documentation is missing, outdated, or was never written. The last update was six years ago; some of the dependencies are abandoned. Whether you call it “tech debt” or “care and feeding” or “maintenance”, it consumes a huge fraction of our time and engineering budget—without a clear path forward. This is the reality many companies face—whether they’re in finance , healthcare, or the public sector—they have no choice but to pour millions of dollars into just keeping these systems afloat. Or do they? At MongoDB, we're building a new kind of engineering capability—one that combines the latest advancements in generative AI with the principles of forward-deployed engineering to help modernize legacy systems at a speed and scale that feels impossible. What is forward-deployed AI engineering? Because it’s new, you may not have heard of this role before. It's a bridge between engineering, consulting, and product development. As an AI Forward Deployed Engineer, you won’t just sit behind the scenes writing code. You'll be embedded with our customers, working side by side with their teams to solve real-world modernization challenges. No theory, just hands-on engineering at the sharpest edge of AI innovation. Your goal? Write software with AI, write software for AI, at speeds you’ve never experienced. In some tasks, we have benchmarked AI as being more than eight hundred times faster than a human being… working in that environment is, I assure you, radically different. You’ll deliver immediate, meaningful impact—whether that’s untangling a million-line code base, modernizing outdated Java frameworks, or helping teams migrate from niche, unsupported languages to modern tech stacks. MongoDB + AI: Changing the game One of the most exciting parts of our work is how AI is fundamentally changing what's possible. In the past, a modernization project like this might take five years and dozens of engineers. Today, with the power of MongoDB and AI, that same work is done in a fraction of the time—sometimes in months, sometimes in weeks—with a much smaller, highly focused team. We use LLMs and other AI tools to do things like: Add missing documentation to legacy code Write unit tests where none existed Remove outdated frameworks, replacing them with others Analyze and map massive, messy code bases Move between programming languages, frameworks, platforms, ORMs, databases, and front-end technologies. Let me be clear—none of this is easy, or on autopilot. We augment great engineers with powerful tools so they can focus on the work that matters most, so they can think big, and go far . Combining human expertise with AI capability leads to outcomes no one thought possible. Why it matters Modernizing these systems isn’t just about efficiency or cost savings—it’s mission-critical. The legacy platforms power trading systems, healthcare infrastructure, and governmental services that support real people every day. In multiple countries around the world. Being able to show them a fully functional prototype in weeks, instead of years, is game-changing. It proves what's possible. It builds momentum. It allows them to rethink how they, too, can structure teams and processes in a world where technology moves faster than ever. Why join MongoDB? If you're an engineer who thrives on autonomy, problem-solving, and building real solutions with immediate impact—and you want to work with AI every day in novel and complex ways— this is your chance. We’re looking for folks from a variety of backgrounds—software development, consulting, product engineering, or technical architecture—who are passionate about learning and applying new technologies, skilled communicators who are excited to partner with customers, and who are comfortable operating in ambiguity. As part of our Application Modernization function, you will: Work directly with customers and stakeholders Rapidly build solutions that solve meaningful business challenges Operate at the intersection of engineering, product, and consulting Learn and apply cutting-edge AI, inventing both new methodologies and technologies Be part of a small, fast-moving, high-impact team You might even write a white paper or two. You’ll also be part of a culture that values leadership at every level. At MongoDB, we believe in making it matter —and this role is designed to make change, not just for our customers, but for the software industry as a whole. If you’ve ever wanted to be part of a startup within an enterprise, this is your opportunity. We're building a new way of working — one where experimentation, agility, and ownership are at the core. Help us build what’s next At MongoDB, we’re not just modernizing applications—we’re modernizing how software is made . If that excites you, if you want to shape the future of software development alongside a team of builders, thinkers, and innovators—we’d love to hear from you. Check out our open roles and join us in redefining what's possible.

April 23, 2025

Advancing Integration Between Drupal and MongoDB

MongoDB regularly collaborates with open source innovators like David Bekker, a Drupal core contributor with over 600 commit credits. David's expertise lies in Drupal's Database API and database driver modules, and he's passionate about delivering business value through open source development. Drupal is a widely used open-source content management system known for its robustness and flexibility, enabling users to create everything from personal blogs to enterprise-level applications. While Drupal typically relies on relational databases (e.g.,MySQL), there has been growing interest in the Drupal community in exploring how modern databases like MongoDB can improve efficiency. In this guest post, David explores integrating MongoDB with Drupal to enhance its performance and scalability, helping Drupal remain competitive in the digital landscape. - Rishabh Bisht, Product Manager, Developer Experience Who am I? Hello! My name is David Bekker (a.k.a. daffie ), and I’m a seasoned Drupal core contributor with over 600 commit credits. I maintain Drupal’s Database API and database driver modules. My passion lies in open source development, driven by a desire to create maximum business value. When I was looking for a new high-impact project to work on, I chose to develop a MongoDB driver for Drupal —one that stores entity instances as JSON objects. This project addresses Drupal’s evolving needs in a meaningful way. User-centric innovation: Drupal’s next evolution Drupal is rapidly evolving, making it particularly suitable for community and client portal solutions. This progression introduces new technical requirements, especially for authenticated, session-based scenarios like intranets and dashboards, which benefit from more adaptable storage solutions. While Drupal's abstract database layer remains tied to the relational models, embracing NoSQL databases would better support its evolving needs for modern applications. To understand why this shift is crucial, let's compare this transition to a challenge Drupal faced years ago: optimizing sites for mobile devices. Back then, significant changes were needed to enhance mobile usability. Now, we face a similar paradigm shift as the market evolves from sites for anonymous users to those centered on authenticated users. Drupal must adapt, and Drupal on MongoDB is the key to this transformation. MongoDB, with its flexible, JSON-based structure, complements Drupal's architecture well. A robust integration with MongoDB would enhance capabilities and better equip Drupal to meet the expanding demands of enterprises. Beyond traditional use cases, Drupal on MongoDB is also ideal as a backend for iOS, Android, and JavaScript applications, providing personalized and scalable solutions. Redefining data storage and retrieval Drupal on MongoDB is more than just a new database option. It enhances Drupal’s ability to compete in a changing digital landscape. Drupal’s robust entity system provides a solid foundation where everything is structured as an entity. Traditionally, Drupal leverages relational databases like MySQL or MariaDB, efficiently managing data across multiple tables. This approach performs well for sites with a large number of anonymous users. However, for sites with many authenticated users, the complexity of retrieving entity data from multiple tables can introduce performance challenges. Optimizing data retrieval can significantly enhance the user experience, making Drupal even more powerful for dynamic, user-centric applications. With MongoDB, every Drupal entity instance is stored as a single JSON object, including all revisions, translations, and field data. This streamlined data structure allows for significantly faster retrieval, making Drupal a stronger solution for personalized, user-focused experiences. As the market shifts toward authentication-driven sites, supporting MongoDB ensures that Drupal remains a competitive and scalable option. Rather than replacing Drupal’s strengths, this integration enhances them, allowing Drupal to meet modern performance demands while maintaining its flexibility and power. Scalability: Why MongoDB makes sense for large Drupal projects The scalability of NoSQL databases like MongoDB sets them apart from traditional relational databases such as MySQL or MariaDB. While relational databases typically rely on a single-server model, MongoDB supports horizontal scaling, enabling distributed setups with thousands of servers acting as a unified database. This architecture provides the performance needed for large-scale projects with millions of authenticated users. As community-driven software, Drupal is built to support interactive, user-focused experiences, including forums, profiles, and content management. Traditionally, its relational model organizes data across multiple tables—similar to storing the chapters of a book separately in a library. This approach ensures data consistency and flexibility, making it highly effective for managing structured content. However, as the demand for authentication-heavy sites grows, the way data is stored becomes a crucial factor in performance. MongoDB offers a more efficient alternative by storing entire entities as JSON objects—just like keeping an entire book intact rather than splitting it into separate chapters across different locations. This eliminates the need for complex table joins, significantly accelerating data retrieval and making MongoDB well suited for personalized dashboards and dynamic content feeds. For small-scale sites, both relational and NoSQL approaches work. But when scalability, speed, and efficiency become priorities—particularly for sites with millions of authenticated users—MongoDB provides a natural and powerful solution for taking Drupal to the next level. Example of a user entity stored in MongoDB The sample document below is an example of how a user entity could look like in MongoDB, containing fields like _id , uid , uuid , and langcode . It includes an embedded user_translations array that holds user details such as name , email , timezone , status , and timestamps for various activities. { _id: ObjectId('664afdd4a3a001e71e0b49c7'), uid: 1, uuid: '841149cd-fe56-47c4-a112-6d23f561332f', langcode: 'en', user_translations: [ { uid: 1, uuid: '841149cd-fe56-47c4-a112-6d23f561332f', langcode: 'en', preferred_langcode: 'en', name: 'root', pass: '$2y$10$kjGuIsPOTDa2TseuWMFGS.veLzH/khl0SfsuZNAeRPRtABgfq5GSC', mail: 'admin@example.com', timezone: 'Europe/Amsterdam', status: true, created: ISODate('2024-05-20T07:37:54.000Z'), changed: ISODate('2024-05-20T07:42:08.000Z'), access: ISODate('2024-05-20T08:46:47.000Z'), login: ISODate('2024-05-20T07:44:16.000Z'), init: 'admin@example.com', default_langcode: true, user_translations__roles: [ { bundle: 'user', deleted: false, langcode: 'en', entity_id: 1, revision_id: 1, delta: 0, roles_target_id: 'administrator' } ] } ], login: ISODate('2024-05-20T07:44:16.000Z'), access: ISODate('2024-05-20T08:46:47.000Z') } Optimizing data storage for performance Switching to MongoDB alone isn’t enough to make Drupal a top-tier solution for sites with a high number of authenticated users. Indeed, developers must rethink how data is stored. In traditional Drupal setups optimized for anonymous users, caching mechanisms like Redis compensate for slow database queries. However, for authenticated users, where content is dynamic and personalized, this approach falls short. Drupal itself needs to be fast, not just its caching layer. MongoDB enables developers to store data in the way the application uses it, reducing the need for complex queries that slow down performance. Instead of relying on costly operations like joins and subqueries, simple and efficient queries should be the norm. Tools like materialized views—precomputed query results stored as database tables—help achieve this, ensuring faster data retrieval while keeping the database structured for high performance. Why MongoDB for Drupal? While many databases support JSON storage, MongoDB is the only one that fully meets Drupal’s needs . Its capabilities extend beyond basic JSON support, making it the optimal choice for storing entity instances efficiently. Additionally, MongoDB offers several key advantages that align with Drupal’s evolving requirements: Horizontal scaling: Easily distribute database load across multiple servers, making it scalable for large user bases. Integrated file storage: Store user-uploaded files directly in the database instead of on the web server, simplifying hosting. Built-in full-text search: Eliminates the need for separate search solutions like SOLR, reducing infrastructure complexity. AI capabilities: Supports AI vectors, allowing for features like advanced search and personalization tailored to a site’s content. Current status Drupal’s journey to embracing more flexible data storage solutions is advancing with promising developments: The MongoDB driver for Drupal is available as a contrib module for Drupal 11, with over 99% of core tests passing. Discussions are ongoing to merge MongoDB support into Drupal core, pending community contributions. Finalist / Tech Blog is already running entirely on MongoDB. These steps mark a significant transition for Drupal, showcasing its evolution towards accommodating non-relational databases like MongoDB. It paves the way for broader applications and more robust infrastructure by leveraging MongoDB’s strengths in flexibility and scalability. Conclusion As the web moves toward more personalized, user-centric experiences, Drupal must evolve to remain competitive. MongoDB is a key enabler of this evolution, providing faster, more scalable solutions for authenticated user-heavy sites. By embracing MongoDB, Drupal developers can unlock new performance possibilities, simplify infrastructure, and build future-ready web applications. Check out the tutorial on how to run Drupal on MongoDB Atlas and start experiencing the benefits of this powerful integration today! Want to get involved? Join the conversation in the Drupal community via Slack in the #mongodb and #contribute channels. Let’s shape the future of Drupal together!

April 22, 2025

Transforming News Into Audio Experiences with MongoDB and AI

You wake up, brew your coffee, and start your day with a perfectly tailored podcast summarizing the latest news—delivered in a natural, engaging voice. No manual curation, no human narration, just seamless AI magic. Sounds like the future? It's happening now, powered by MongoDB and generative AI. In 2025, the demand for audio content—particularly podcasts—surged, with 9 million new active listeners in the United States alone, prompting news organizations to seek efficient ways to deliver daily summaries to their audiences. However, automating news delivery has proven to be a challenging task, as media outlets must manage dynamic article data and convert this information into high-quality audio formats at scale. To overcome these hurdles, media organizations can use MongoDB for data storage alongside generative AI for podcast creation, developing a scalable solution for automated news broadcasting. This approach unlocks new AI-driven business opportunities and can attract new customers while strengthening the loyalty of existing ones, contributing to increased revenue streams for media outlets. Check out our AI Learning Hub to learn more about building AI-powered apps with MongoDB. The secret sauce: MongoDB + AI In a news automation solution, MongoDB acts as the system’s backbone, storing news article information as flexible documents with fields like title, content, and publication date—all within a single collection. Alongside this, dynamic elements (such as the number of qualified reads) can be seamlessly integrated into the same document to track content popularity. Moreover, derived insights—e.g., sentiment analysis and key entities—can be generated and enriched through a gen AI pipeline directly within the existing collection. Figure 1. MongoDB data storage for media. This adaptable data structure ensures that the system remains both efficient and scalable, regardless of content diversity or evolving features. As a result, media outlets have created a robust framework to query and extract the latest news and metadata from MongoDB. They can now integrate AI with advanced language models to transform this information into an audio podcast. With this foundation in place, let's examine why MongoDB is well-suited for implementing AI-driven applications. Why MongoDB is the perfect fit News data is inherently diverse, with each article containing a unique mix of attributes, including main content fields (e.g. id, title, body, date, imageURL), calculated meta data (e.g. read count), generated fields with the help of GenAI (e.g. keywords, sentiment) and embeddings for semantic/vector search. Some of these elements originate from publishers, while others emerge from user interactions or AI-driven analysis. MongoDB’s flexible document model accommodates all these attributes—whether predefined or dynamically generated, within a single, adaptable structure. This eliminates the rigidity of traditional databases and ensures that the system evolves seamlessly alongside the data it manages. What’s more, speed is critical in news automation. By storing complete, self-contained documents, MongoDB enables rapid retrieval and processing without the need for complex joins. This efficiency allows articles to be enriched, analyzed, and transformed into audio content in near real-time. And scalability is built in. Whether handling a small stream of updates or processing vast amounts of constantly changing data, MongoDB’s distributed architecture ensures high availability and seamless growth, making it ideal for large-scale media applications. Last but hardly least, developers benefit from MongoDB’s agility. Without the constraints of fixed schemas, new data points—whether from evolving AI models, audience engagement metrics, or editorial enhancements—can be integrated effortlessly. This flexibility allows teams to experiment, iterate, and scale without friction, ensuring that the system remains future-proof as news consumption evolves. Figure 2. MongoDB benefits for AI-driven applications. Bringing news to life with generative AI Selecting MongoDB for database storage is just the beginning; the real magic unfolds when text meets AI-powered speech synthesis. In our labs, we have experimented with Google’s NotebookLM model to refine news text, ensuring smooth narration with accurate intonation and pacing. Putting all these pieces together, the diagram below illustrates the workflow for automating AI-based news summaries into audio conversions. Figure 3. AI-based text-to-audio conversion architecture. The process begins with a script that retrieves relevant news articles from MongoDB, using the Aggregation Framework and Vector Search to ensure semantic relevance. These selected articles are then passed through an AI-powered pipeline, where they are condensed into a structured podcast script featuring multiple voices. Once the script is refined, advanced text-to-speech models transform it into high-quality audio, which is stored as a .wav file. To optimize delivery, the generated podcast is cached, ensuring seamless playback for users on demand. The result? A polished, human-like narration, ready for listeners in MP3 format. Thanks to this implementation, media outlets can finally let go of the robotic voices of past automations. Instead, they can now deliver a listening experience to their customers that's human, engaging, and professional. The future of AI-powered news consumption This system isn’t just a technological innovation; it’s a revolution in how we consume news. By combining MongoDB’s efficiency with AI’s creative capabilities, media organizations can deliver personalized, real-time news summaries without human intervention. It’s faster, smarter, and scalable—ushering in a new era of automated audio content. Want to build the next-gen AI-powered media platform? Start with MongoDB and let your content speak for itself! To learn more about integrating AI into media systems using MongoDB, check out the following resources to guide your next steps: The MongoDB Solutions Library: Gen AI-powered video summarization The MongoDB Blog: AI-Powered Media Personalization: MongoDB and Vector Search

April 21, 2025

Away From the Keyboard: Kyle Lai, Software Engineer 2

In “Away From the Keyboard,” MongoDB developers discuss what they do, how they keep a healthy work-life balance, and their advice for people seeking a more holistic approach to coding. In this article, Kyle Lai describes his role as a Software Engineer 2 at MongoDB; why he’d rather not be like the characters on the TV show, Severance; and how his commute helps set boundaries between his professional and personal lives. Q: What do you do at MongoDB? Kyle: Hi! I’m an engineer on Atlas Growth 1, where we run experiments on Atlas and coordinate closely with marketing, product, design, and analytics to improve the user experience. Atlas Growth 1 is part of the broader Atlas Growth engineering teams, where we own the experimentation platform and experiment software development kit, allowing other teams to run experiments as well! The engineers on Atlas Growth are very involved with the product side of our experiments. We help the analytics team collect metrics and decide if a given experiment was a win. Sometimes, seemingly obvious positive improvements can turn out to be detrimental to the user flow, so our experimentation process allows us to learn greatly about our users, whether the experiment wins or not. Q: What does work-life balance look like for you? Kyle: Work-life balance for me means that I won’t be worrying about responding to messages or needing to open my laptop after work hours. It also means that my teammates equally respect my work-life balance and do not expect me to work during non-work hours. Q: How do you ensure you set boundaries between work and personal life? Kyle: Generally, for me, it’s more difficult to set boundaries between work and personal life when I’m working from home, so I try to come into the office most days. My commute also provides me with time to wind down and signal that work is over for the day. In a way, the drive to and from the train station allows me to transition to getting into the mindset for work or to decompress at the end of the day. Q: Has work-life balance always been a priority for you, or did you develop it later in your career? Kyle: As someone who is early in my career, work-life balance is something that I’ve grown to appreciate and see as a priority in my life. It allows me to enjoy my personal life, and definitely contributes to a healthier me and a healthier team. Q: What benefits has this balance given you in your career? Kyle: Our team has a weekly Friday hangout meeting, where we have a different question posed to us each week. One of the questions was based on the TV show, Severance. Would we choose to be “severed” like the characters in the show? They undergo a procedure that separates their work and personal brains—their work brains have no awareness of their personal lives, and vice versa. As someone who hasn’t seen the show, but has heard about it from the rest of my team, I wouldn’t do it. Balancing my work and personal lives allows me to enjoy each side more. I’m motivated for the end of the week so I can enjoy the weekend, and I’m also excited to come to work with a fresh headspace on Mondays, since I am not overworking during non-work hours. Q: What advice would you give to someone seeking to find a better balance? Kyle: I’ll sometimes have the urge to continue working past work hours, as I’ll feel like I’m about to finish whatever task I’m working on very soon or think I can get even more done if I don’t stop working. That backfires pretty quickly. You have to realize you can be easily fatigued and are not able to give your best work if you constantly keep working. A proper work-life balance will allow you to have a fresh start and a clear mind each day. As for how to better separate work and personal life, I’d suggest changing notification settings on your phone for Slack, so that non-urgent work messages won’t tempt you to open your laptop. Another strategy would be to associate some event with a cutoff for checking work things, such as not reading messages once you’ve left the office or boarded the train. I’ve had teammates tell me they delete Slack from their phones when they’re on vacation, which is a good idea! Thank you to Kyle Lai for sharing these insights! And thanks to all of you for reading. For past articles in this series, check out our interviews with: Staff Engineer, Ariel Hou Senior AI Developer Advocate, Apoorva Joshi Developer Advocate, Anaiya Raisinghani Senior Partner Marketing Manager, Rafa Liou Staff Software Engineer, Everton Agner Interested in learning more about or connecting more with MongoDB? Join our MongoDB Community to meet other community members, hear about inspiring topics, and receive the latest MongoDB news and events. And let us know if you have any questions for our future guests when it comes to building a better work-life balance as developers. Tag us on social media: @/mongodb #LoveYourDevelopers #AwayFromTheKeyboard

April 17, 2025

Unlocking BI Potential with DataGenie & MongoDB

Business intelligence (BI) plays a pivotal role in strategic decision-making. Enterprises collect massive amounts of data yet struggle to convert it into actionable insights. Conventional BI is reactive, constrained by predefined dashboards, and human-dependent, thus making it error-prone and non-scalable. Businesses today are data-rich but insight-poor. Enter DataGenie, powered by MongoDB—BI reimagined for the modern enterprise. DataGenie autonomously tracks millions of metrics across the entire business datascape. It learns complex trends like seasonality, discovers correlations & causations, detects issues & opportunities, connects the dots across related items, and delivers 5 to 10 prioritized actionable insights as stories in natural language to non-data-savvy business users. This enables business leaders to make bold, data-backed decisions without the need for manual data analysis. With advanced natural language capabilities through Talk to Data, users can query their data conversationally, making analytics truly accessible. The challenges: Why DataGenie needed a change DataGenie processes large volumes of enterprise data on a daily basis for customers, tracking billions of time series metrics and performing anomaly detection autonomously to generate deep, connected insights for business users. The below diagram represents the functional layers of DataGenie. Figure 1. DataGenie’s functional layers. Central to the capability of DataGenie is the metrics store, which stores, rolls up, and serves billions of metrics. At DataGenie, we were using an RDBMS (PostgreSQL) as the metrics store. As we scaled to larger enterprise customers, DataGenie processed significantly higher volumes of data. The complex feature sets we were building also required enormous flexibility and low latency in how we store & retrieve our metrics. DataGenie had multiple components that served different purposes, and all of these had to be scaled independently to meet our sub-second latency requirements. With PostgreSQL as the metrics store for quite some time and tried to squeeze it to the maximum extent possible at the cost of flexibility. Since we over-optimized the structure for performance, we lost the flexibility we required to build our next-gen features, which were extremely demanding We defaulted to PostgreSQL for storing the insights (i.e. stories), again optimized for storage and speed, hurting us on the flexibility part For the vector store, we had been using ChromaDB for storing all our vector embeddings. As the data volumes grew, the most challenging part was maintaining the data sync We had to use a different data store for knowledge store and yet another technology for caching The major problems we had were as follows: Rigid schema that hindered flexibility for evolving data needs. High latency & processing cost due to extensive preprocessing to achieve the desired structure Slow development cycles that hampered rapid innovation How MongoDB gave DataGenie a superpower After extensive experimentation with time-series databases, document databases, and vector stores, we realized that MongoDB would be the perfect fit for us since it exactly solved all our requirements with a single database. Figure 2. MongoDB data store architecture. Metrics store When we migrated to MongoDB, we achieved a remarkable reduction in query latency. Previously, complex queries on 120 million documents took around 3 seconds to execute. With MongoDB's efficient architecture, we brought this down to an impressive 350-500 milliseconds for 500M+ docs , representing an 85-90% improvement in query speed for a much larger scale. Additionally, for storing metrics, we transitioned to a key-value pair schema in MongoDB. This change allowed us to reduce our data volume significantly— from 300 million documents to just 10 million documents —thanks to MongoDB's flexible schema and optimized storage. This optimization not only reduced our storage footprint for metrics but also enhanced query efficiency. Insights store By leveraging MongoDB for the insight service, we eliminated the need for extensive post-processing, which previously consumed substantial computational resources. This resulted in a significant cost advantage, reducing our Spark processing costs by 90% or more (from $80 to $8 per job). Querying 10,000+ insights took a minute before. With MongoDB, the same task is now completed in under 6 seconds—a 10x improvement in performance . MongoDB’s flexible aggregation pipeline was instrumental in achieving these results. For example, we extensively use dynamic filter presets to control which insights are shown to which users, based on their role & authority. The MongoDB aggregation pipeline dynamically adapts to user configurations, retrieving only the data that’s relevant. LLM service & vector store The Genie+ feature in DataGenie is our LLM-powered application that unifies all DataGenie features through a conversational interface. We leverage MongoDB as a vector database to store KPI details, dimensions, and dimension values. Each vector document embeds essential metadata, facilitating fast and accurate retrieval for LLM-based queries. By serving as the vector store for DataGenie, MongoDB enables efficient semantic search, allowing the LLM to retrieve contextual, relevant KPIs, dimensions, and values with minimal latency, enhancing the accuracy and responsiveness of Genie+ interactions. Additionally, integrating MongoDB Atlas Search for semantic search significantly improved performance. It provided faster, more relevant results while minimizing integration challenges.MongoDB’s schema-less design and scalable architecture also streamlined data management. Knowledge store & cache MongoDB’s schema-less design enables us to store complex, dynamic relationships and scale them with ease. We also shifted to using MongoDB as our caching layer. Previously, having separate data stores made syncing and maintenance cumbersome. Centralizing this information in MongoDB simplified operations, enabled automatic syncing, and ensured consistent data availability across all features. With MongoDB, DataGenie is reducing time-to-market for feature releases Although we started the MongoDB migration to solve only our existing scalability and latency issues, we soon realized that just by migrating to MongoDB, we could imagine even bigger and more demanding features without engineering limitations. Figure 3. MongoDB + DataGenie integration. DataGenie engineering team refers v2 magic moment since migrating to MongoDB makes it a lot easier & flexible to roll out the following new features: DataGenie Nirvana: A delay in the supply chain for a raw material can cascade into a revenue impact. Conventional analytics relies on complex ETL pipelines and data marts to unify disparate data and deliver connected dashboard metrics. DataGenie Nirvana eliminates the need for a centralized data lake by independently generating aggregate metrics from each source and applying advanced correlation and causation algorithms on aggregated data to detect hidden connections. DataGenie Wisdom: Wisdom leverages an agentic framework & knowledge stores, to achieve two outcomes: Guided onboarding: Onboarding a new use case in DataGenie is as simple as explaining the business problem, success criteria, and sharing sample data - DataGenie autonomously configures itself for relevant metrics tracking to deliver the desired outcome. Next best action: DataGenie autonomously surfaces insights - like a 10% brand adoption spike in a specific market and customer demographics. By leveraging enterprise knowledge bases and domain-specific learning, DataGenie would propose targeted marketing campaigns as the Next Best Action for this insight. Powered by Genie: DataGenie offers powerful augmented analytics that can be quickly configured for any use case and integrated through secure, high-performance APIs. This powers data products in multiple verticals, including Healthcare & FinOps, to deliver compelling augmented analytics as a premium add-on, drastically reducing their engineering burden and GTM risk. All of these advanced features require enormous schema flexibility, low latency aggregation, and a vector database that’s always in sync with the metrics & insights. That’s exactly what we get with MongoDB! Powered by MongoDB Atlas, DataGenie delivers actionable insights to enterprises, helping them unlock new revenue potential and reduce costs. The following are some of the DataGenie use cases in Retail: Demand shifts & forecasting: Proactively adjust inventory or revise marketing strategies based on product demand changes. Promotional effectiveness: Optimize marketing spend by understanding which promotions resonate with which customer segments. Customer segmentation & personalization: Personalize offers based on customer behavior and demographics. Supply chain & logistics: Minimize disruptions by identifying potential bottlenecks and proposing alternative solutions. Inventory optimization: Streamline inventory management by flagging potential stockouts or overstock. Fraud & loss prevention: Detect anomalies in transaction data that may signal fraud or errors. Customer retention & loyalty: Propose retention strategies to address customer churn. Staffing optimization: Optimize customer support staffing. Final thoughts Migrating to MongoDB did more than just solve DataGenie’s scalability and latency challenges - it unlocked new possibilities. The flexibility of MongoDB allowed DataGenie to innovate faster and conceptualize new features such as Nirvana, Wisdom, and ultra-efficient microservices. This transformation stands as a proof of concept for future product companies considering partnering with MongoDB. The partnership between DataGenie and MongoDB is a testament to how the right technology choices can drive massive business value, improving performance, scalability, and cost-efficiency. Ready to unlock deeper retail insights? Head over to our retail page to learn more. Check out our Atlas Learning Hub to boost your MongoDB skills.

April 16, 2025

Introducing Database Digest: Building Foundations for Success

Today at MongoDB .local Toronto , I’m excited to share the first issue of Database Digest —MongoDB’s new digital magazine that explores the critical role of data in powering modern applications. This inaugural issue explores modern data architecture, and shows how—when the right data foundation meets emerging technologies—pioneering companies are fundamentally reimagining what's possible. The dawn of data ubiquity Currently, we stand in what McKinsey calls the " data ubiquity era "— with data flowing through organizations as essentially as electricity powers the modern world. The transformation to this era has brought both unprecedented opportunity and formidable challenges. Organizations must simultaneously manage huge volumes of data while delivering the real-time, personalized experiences that define competitive advantage today. Successfully doing so doesn’t mean merely adopting new technologies. Instead, it requires fundamentally rethinking how data is stored, processed, and leveraged to drive business value. Traditional relational database systems simply cannot meet these demands. The future belongs to organizations with data architectures designed for the agility, scalability, and versatility needed to handle diverse data types while seamlessly integrating with emerging technologies like AI. The cornerstone of AI success The rise of AI, and the speed at which the market has been changing have fundamentally shifted the importance of adaptability. However, software can only adapt as fast as its foundation allows. At MongoDB, we believe modern databases are the cornerstone of the age of AI, providing the essential capabilities needed for success in this new era. To do so, they must be able to: Handle all forms of data and provide intelligent search: Modern databases consolidate structured and unstructured data into a single system, eliminating silos that restrict AI innovation. They ground AI output in accurate, contextual data that drives better outcomes. Scale without constraints and react instantly: Databases should be able to adapt to unpredictable workloads and massive data volumes without performance degradation. They should enable real-time decisions and actions when opportunities or threats emerge. Embed domain-specific AI and secure data throughout: Modern databases enhance accuracy with specialized models that reduce hallucinations, and they protect information at every stage without sacrificing speed or functionality. Figure 1. Modern database demands. The impact of a modern database on AI innovation isn’t theoretical—we're seeing organizations like Swisscom leverage this approach to apply generative AI to their extensive expert content library, transforming how they serve the banking industry by delivering bespoke, contextual information within seconds. The AI revolution Perhaps nowhere is the impact of a modern data foundation more profound than in the rapid evolution of AI applications. In just a short time, we've progressed from simple LLM-powered chatbots to more advanced agentic systems capable of understanding complex goals, breaking them into manageable steps, and executing them autonomously—all while maintaining context and learning from interactions. This represents more than incremental progress—it's a fundamental shift in how AI serves as a strategic partner in solving business challenges. MongoDB sits at the heart of this transformation, providing the critical bridge between AI models and data while enabling vector storage, real-time processing, and seamless integration with LLM orchestrators. Companies like Questflow demonstrate the power of this approach, revolutionizing the future of work through a decentralized, autonomous AI agent network that orchestrates multiple AI agents collaborating with humans. By leveraging MongoDB's flexible document model and vector search capabilities, they're enabling startups to create dynamic AI solutions that handle everything from data analysis to content creation, while maintaining context and learning from interactions. Modernizing legacy systems: the strategic imperative For established enterprises, the journey to a modern data foundation often begins with addressing the legacy systems that consume up to 80% of IT budgets yet constrain innovation. Modernization isn't just a technical upgrade—it's a strategic move toward growth, efficiency, and competitive advantage. The evidence is compelling: Bendigo and Adelaide Bank achieved a staggering 90% reduction in both time and cost modernizing core banking applications using MongoDB's repeatable modernization framework and AI-powered migration tools. This transformation isn't just about cost savings—it's about creating the foundation for entirely new capabilities that drive business value. Modern data architecture must embody flexibility across multiple dimensions—from supporting diverse data models to providing deployment options spanning cloud-native, cloud-agnostic, and on-premise environments. This approach enables organizations to break free from silos, integrate AI capabilities, and create a unified data foundation supporting both current operations and future innovations. What’s next for data The organizations featured throughout Database Digest share a common vision: they recognize that tomorrow's success depends on today's data foundation. The convergence of flexible document models, advanced AI integration, and cloud-native capabilities isn't just enabling incremental improvements—it's powering applications and experiences that were never before possible. So I invite you to explore the full, inaugural issue of Database Digest to discover how MongoDB is helping organizations across industries build the foundation for tomorrow's success. This isn't just about technology—it's about creating the foundation for transformation that delivers real business value in our increasingly data-driven world. Visit mongodb.com/library/database-digest to download your copy and join us on this journey into the future of data.

April 15, 2025

Now Generally Available: 7 New Resource Policies to Strengthen Atlas Security

Organizations demand for a scalable means to enforce security and governance controls across their database deployments without slowing down developer productivity. To address this, MongoDB introduced resource policies in public preview on February 10th, 2025. Resource policies enable organization administrators to set up automated, organization-wide ‘guardrails’ for their MongoDB Atlas deployments. At public preview, three policies were released to this end. Today, MongoDB is announcing the general availability of resource policies in MongoDB Atlas. This release introduces seven additional policies and a new graphical user interface (GUI) for creating and managing policies. These enhancements give organizations greater control over MongoDB Atlas configurations, simplifying security and compliance automation. How resource policies enable secure innovation Innovation is essential for organizations to maintain competitiveness in a rapidly evolving global landscape. Companies with higher levels of innovation outperformed their peers financially, according to a Cornell University study analyzing S&P 500 companies between 1998 and 2023 1 . One of the most effective ways to drive innovation is by equipping developers with the right tools and giving them the autonomy to put them into action 2 . However, without standardized controls governing those tools, developers can inadvertently configure Atlas clusters to deviate from corporate or regulatory best practices. Manual approval processes for every new project create delays. Concurrently, platform teams struggle to enforce consistent security policies across the organization, leading to increased complexity and costs. As cybersecurity threats evolve daily and regulations tighten, granting developers autonomy and quickly provisioning access to essential tools can introduce risks. Organizations must implement strong security measures to maintain compliance and enable secure innovation. Resource policies empower organizations to enforce security and compliance standards across their entire Atlas environment. Instead of targeting specific user groups, these policies establish organization-wide guardrails to govern how Atlas can be configured. This reduces the risk of misconfigurations and security gaps. With resource policies, security and compliance standards are applied automatically across all Atlas projects and clusters. This eliminates the need for manual approvals. Developers gain self-service access to the resources they need while remaining within approved organizational boundaries. Simultaneously, platform teams can centrally manage resource policies to ensure consistency and free up time for strategic initiatives. Resource policies strengthen security, streamline operations, and help accelerate innovation by automating guardrails and simplifying governance. Organizations can scale securely while empowering developers to move faster without compromising compliance. What resource policies are available? table, th, td { border: 1px solid black; } Policy Type Description Available Since Restrict cloud provider Ensure clusters are only deployed on approved cloud providers (AWS, Azure, or Google Cloud). This prevents accidental or unauthorized deployments in unapproved environments. This supports organizations in meeting regulatory or business requirements. Public preview

April 14, 2025

GraphRAG with MongoDB Atlas: Integrating Knowledge Graphs with LLMs

A key challenge AI developers face is providing context to large language models (LLMs) to build reliable AI-enhanced applications; retrieval-augmented generation (RAG) is widely used to tackle this challenge. While vector-based RAG, the standard (or baseline) implementation of retrieval-augmented generation, is useful for many use cases, it is limited in providing LLMs with reasoning capabilities that can understand relationships between diverse concepts scattered throughout large knowledge bases. As a result, the accuracy of vector RAG-enhanced LLM outputs in applications can disappoint—and even mislead—end users. Now generally available, MongoDB Atlas ’ new LangChain integration for GraphRAG—a variation of RAG architecture that integrates a knowledge graph with LLMs—can help address these limitations. GraphRAG: Connecting the dots First, a short explanation of knowledge graphs: a knowledge graph is a structured representation of information in which entities (such as people, organizations, or concepts) are connected by relationships. Knowledge graphs work like maps, and show how different pieces of information relate to each other. This structure helps computers understand connections between facts, answer complex questions, and find relevant information more easily. Traditional RAG applications split knowledge data into chunks, vectorize them into embeddings, and then retrieve chunks of data through semantic similarity search; GraphRAG builds on this approach. But instead of treating each document or chunk as an isolated piece of information, GraphRAG considers how different pieces of knowledge are connected and relate to each other through a knowledge graph. Figure 1. Embedding-based vector search vs. entity-based graph search. GraphRAG improves RAG architectures in three ways: First, GraphRAG can improve response accuracy . Integrating knowledge graphs into the retrieval component of RAG has shown significant improvements in multiple publications. For example, benchmarks in the AWS investigation, “ Improving Retrieval Augmented Generation Accuracy with GraphRAG ” demonstrated nearly double the correct answers compared to traditional embedding-based RAG. Also, embedding-based methods rely on numerical vectors and can make it difficult to interpret why certain chunks are related. Conversely, a graph-based approach provides a visual and auditable representation of document relationships. Consequently, GraphRAG offers more explainability and transparency into retrieved information for improved insight into why certain data is being retrieved. These insights can help optimize data retrieval patterns to improve accuracy. Finally, GraphRAG can help answer questions that RAG is not well-suited for—particularly when understanding a knowledge base's structure, hierarchy, and links is essential . Vector-based RAG struggles in these cases because breaking documents into chunks loses the big picture. For example, prompts like “What are the themes covered in the 2025 strategic plan?” are not well handled. This is because the semantic similarity between the prompt, with keywords like “themes,” and the actual themes in the document may be weak, especially if they are scattered across different sections. Another example prompt like, “What is John Doe’s role in ACME’s renewable energy projects?” presents challenges because if the relationships between the person, the company, and the related projects are mentioned in different places, it becomes difficult to provide accurate responses with vector-based RAG. Traditional vector-based RAG can struggle in cases like these because it relies solely on semantic similarity search. The logical connections between different entities—such as contract clauses, legal precedents, financial indicators, and market conditions—are often complex and lack semantic keyword overlap. Making logical connections across entities is often referred to as multi-hop retrieval or reasoning in GraphRAG. However, GraphRAG has its own limitations, and is use-case dependent to achieve better accuracy than vector-based RAG: It introduces an extra step: creating the knowledge graph using LLMs to extract entities and relationships. Maintaining and updating the graph as new data arrives becomes an ongoing operational burden. Unlike vector-based RAG, which requires embedding and indexing—a relatively lightweight and fast process—GraphRAG depends on a large LLM to accurately understand, map complex relationships, and integrate them into the existing graph. The added complexity of graph traversal can lead to response latency and scalability challenges as the knowledge base grows. Latency is closely tied to the depth of traversal and the chosen retrieval strategy, both of which must align with the specific requirements of the application. GraphRAG introduces additional retrieval options . While this allows developers more flexibility in the implementation, it also adds complexity. The additional retrieval options include keyword and entity-based retrieval, semantic similarity on the first node, and more. MongoDB Atlas: A unified database for operational data, vectors, and graphs MongoDB Atlas is perfectly suited as a unified database for documents, vectors, and graphs. As a unified platform, it’s ideal for powering LLM-based applications with vector-based or graph-based RAG. Indeed, adopting MongoDB Atlas eliminates the need for point or bolt-on solutions for vector or graph functionality, which often introduce unnecessary complexity, such as data synchronization challenges that can lead to increased latency and potential errors. The unified approach offered by MongoDB Atlas simplifies the architecture and reduces operational overhead, but most importantly, it greatly simplifies the development experience. In practice, this means you can leverage MongoDB Atlas' document model to store rich application data, use vector indexes for similarity search, and model relationships using document references for graph-like structures. Implementing GraphRAG with MongoDB Atlas and LangChain Starting from version 0.5.0, the langchain-mongodb package introduces a new class to simplify the implementation of a GraphRAG architecture. Figure 2. GraphRAG architecture with MongoDB Atlas and LangChain First, it enables the automatic creation of a knowledge graph. Under the hood, it uses a specific prompt sent to an LLM of your choice to extract entities and relationships, structuring the data to be stored as a graph in MongoDB Atlas. Then, it sends a query to the LLM to extract entities and then searches within the graph to find connected entities, their relationships, and associated data. This information, along with the original query, then goes back to the LLM to generate an accurate final response. MongoDB Atlas’ integration in LangChain for GraphRAG follows an entity-based graph approach. However, you can also develop and implement your own GraphRAG with a hybrid approach using MongoDB drivers and MongoDB Atlas’ rich search and aggregation capabilities. Enhancing knowledge retrieval with GraphRAG GraphRAG complements traditional RAG methods by enabling deeper understanding of complex, hierarchical relationships, supporting effective information aggregation and multi-hop reasoning. Hybrid approaches that combine GraphRAG with embedding-based vector search further enhance knowledge retrieval, making them especially effective for advanced RAG and agentic systems. MongoDB Atlas’ unified database simplifies RAG implementation and its variants, including GraphRAG and other hybrid approaches, by supporting documents, vectors, and graph representations in a unified data model that can seamlessly scale from prototype to production. With robust retrieval capabilities ranging from full-text and semantic search to graph search, MongoDB Atlas provides a comprehensive solution for building AI applications. And its integration with proven developer frameworks like LangChain accelerates the development experience—enabling AI developers to build more advanced and efficient retrieval-augmented generation systems that underpin AI applications. Ready to dive into GraphRAG? Learn how to implement it with MongoDB Atlas and LangChain. Head over to the Atlas Learning Hub to boost your MongoDB skills and knowledge.

April 14, 2025

Driving Retail Loyalty with MongoDB and Cognigy

Retail is one of the fastest moving industries, often the very first to leverage cutting-edge AI to create next-gen experiences for their customers. One of the latest areas we’re seeing retailers invest in is agentic AI: they are creating conversational chatbot “agents” that are pulling real-time information from their systems, using Natural Language processing to create conversational responses to customer queries, and then taking action- completing tasks and solving problems. In this race to stay ahead of their competition, retailers today are struggling to quickly bring to market these agents and don’t always have the AI skills in-house. Many are looking to the broad ecosystem of off-the-shelf solutions to leverage the best of what’s already out there—reducing time to market for their AI agents and leaving the AI models and integrations to the experts in the field. Some of the most successful retail conversational AI agents we’ve seen are built on Cognigy , a global leader in customer service solutions. With Cognigy, retailers are quickly spinning up conversational AI agents on top of their MongoDB data to create personalized conversational experiences that not only meet but anticipate customer expectations. Increasingly, whether or not retailers offer customers immediate, seamless interactions are key to retaining their loyalty. Why next-gen conversational AI matters in retail Customer loyalty has been declining yearly, and customers are moving to retailers who can provide an elevated experience at every interaction. According to HubSpot’s 2024 annual customer service survey , 90% of customers expect an immediate response to their inquiries, highlighting how speed has become a critical factor in customer satisfaction. Additionally, 45.9% of business leaders prioritize improving customer experience over product and pricing , demonstrating that in retail, speed and personalization are no longer optional as they define whether a customer stays or moves on. The chatbots of the past that relied on simple rules-based engines and static data don’t meet these customers' new expectations as they lack real-time business context, and can generate misleading answers as they’re not training on the retailer's in-house data sets. This is where Cognigy’s AI agents can create a more compelling experience: These intelligent systems integrate real-time business data with the capabilities of LLMs, enabling AI-driven experiences that are not only personalized but also precise and controlled. Instead of leaving responses open to interpretation, retailers can customize interactions , guide users through processes, and ensure AI-driven recommendations align with actual inventory, customer history, and business rules. This level of contextual understanding and action creates trust-driven experiences that foster loyalty. Having quality data and the ability to harness it effectively is the only way to meet the strategic imperatives that customers demand today. This requires key factors such as being fast, flexible, and high-performing at the scale of your business operations, as winning companies must store and manage their information efficiently. This is where MongoDB, a general-purpose database, truly shines. It is designed to manage your constantly evolving business data, such as inventory, orders, transaction history, and user preferences. MongoDB’s document model stands out in the retail industry, offering the flexibility and scalability businesses need to thrive in today’s fast-paced environment. Cognigy can use this real-time operational data from MongoDB as a direct input to build, run, and deploy conversational AI agents at scale. With just a few clicks, businesses can create AI-driven chatbots and voice agents powered by large language models (LLMs), following their business workflows in a smooth and easy-to-implement way. These agents can seamlessly engage with customers across various phone lines as a major driver for customer interactions, including website chat, Facebook Messenger, and WhatsApp, offering personalized interactions. On the back end, Cognigy is built on MongoDB as its operational data store, taking full advantage of MongoDB’s scalability and high performance to ensure that its conversational AI systems can efficiently process and store large volumes of real-time data while maintaining high availability and reliability. The power of combining AI agents with real-time business data transforms personalization from a static concept into a dynamic ever-evolving experience that makes customers feel truly recognized and understood at every touchpoint. By harnessing these intelligent systems, retailers can go beyond generic interactions to deliver seamless, relevant, and engaging experiences that naturally strengthen customer relationships. Ultimately, true personalization isn’t just about efficiency; it’s about creating meaningful connections that drive lasting customer engagement and loyalty. Let’s look at how this looks in the Cognigy interface when you’re creating a flow for your chatbot: What’s happening behind the scenes? Figure 1 below shows an example customer journey, and demonstrates how Cognigy and MongoDB work together to use real-time data to give reliable and conversational responses to customer questions: Figure 1. An Agentic AI conversational flow with Cognigy pulling user and order data from MongoDB This user’s journey starts when they make a purchase on a retailer’s ecommerce application. The platform securely stores the order details, including product information, customer data, and order status, in MongoDB. To coordinate the delivery, the user reaches out via a chatbot or phone conversation orchestrated by Cognigy AI agents, using advanced Large Language Models (LLMs) to understand the user’s inquiries and respond in a natural, conversational tone. The AI agent retrieves the necessary user information and order details from MongoDB, configured as the data source, taking real-time data that is always up to date. By understanding the user’s query, the agent retrieves the appropriate database information and is also able to update the database with any relevant information generated during the conversation, such as modifying a delivery appointment. As the user schedules their delivery, Cognigy updates the information directly in MongoDB, leveraging features like triggers and change streams to seamlessly synchronize real-time data with other key systems in the customer journey, such as inventory management and delivery providers. This ensures personalized user experiences at every interaction. Shaping the future of customer service with MongoDB and Cognigy Delivering responsive, personalized customer service is more essential than ever. By combining MongoDB’s flexible, versatile, and performant data management with Cognigy’s powerful conversational AI, businesses can create seamless, real-time interactions that keep customers engaged. The future of customer service is fast, dynamic, and seamlessly integrated into business operations. With MongoDB and Cognigy, organizations can harness the power of AI to automate and personalize customer interactions in real time, without the need for extensive development efforts. The MongoDB-Cognigy integration enables businesses to scale context-driven interactions, strengthen customer relationships, and exceed expectations while building lasting customer loyalty. Learn more about how Cognigy built a leading conversational AI solution with MongoDB on our customer story page. Needing a solution for your retail needs? Head over to our retail solutions page to learn how MongoDB supports retail innovation. Read our blog to learn how to enhance retail solutions with retrieval-augmented generation (RAG).

April 10, 2025

What’s New From MongoDB at Google Cloud Next 2025

At Google Cloud Next '25, MongoDB is excited to celebrate a deepening collaboration with Google Cloud, focused on delivering cutting-edge solutions that empower developers, enterprises, and startups alike. The event comes as MongoDB Atlas adds availability for Google Cloud regions in Mexico and South Africa, further expanding joint customers’ ability to deploy, scale, and manage their applications closer to their users while meeting local compliance and performance requirements. MongoDB is also honored to have achieved the 2025 Google Cloud Partner of the Year for Data & Analytics - Marketplace. This award is a testament to the enterprise-scale success stories driven by our combined data and analytics solutions. It’s also MongoDB’s sixth consecutive year as a Google Cloud Partner of the Year, reflecting the relentless innovation and customer-first mindset that define MongoDB’s partnership with Google Cloud. This is in addition to achieving the Google Cloud Ready – Regulated and Sovereignty Solutions Badge. The designation is a major milestone for MongoDB, and demonstrates our ability to deliver compliant and secure solutions that meet the highest standards for data sovereignty. More broadly, we’ve been focused on expanding our collaboration in order to unlock new opportunities for customers to enhance developer productivity, launch AI-powered applications, and do more with their data in 2025. Read on to learn more about what we’ve been working on. Enhancing developer productivity with gen AI For over a decade, MongoDB and Google Cloud have established a rich track record of making it easier, faster, and more secure to build enterprise-grade applications. Our latest gen AI collaborations further this mission, simultaneously enhancing innovation and efficiency. MongoDB is proud to be a launch partner for Google Cloud’s Gemini Code Assist. Announced in December and launching for MongoDB users this week at Google Cloud Next, our integration with Gemini Code Assist enables developers to seamlessly access the latest MongoDB documentation and code snippets within their IDEs. This innovative integration enhances developer productivity by providing immediate access to MongoDB resources, making development workflows more efficient by keeping developers 'in the flow' rather than having to hop in and out of the IDEs to find the information and code examples they need. MongoDB is also expanding our presence in Project IDX , an AI-assisted development workspace for full-stack, multiplatform applications. With MongoDB templates now available in IDX, developers can quickly set up MongoDB environments without leaving their IDE, accelerating the development of generative AI applications and other cloud-based solutions. Learn more by reading this blog post from Google. Developers building applications in Firebase can now integrate MongoDB Atlas with a few clicks. The new Firebase extension for MongoDB Atlas eliminates the need for complex query pipelines or manual data transfers, making it easier than ever to deploy and scale apps leveraging MongoDB Atlas as a vector database in Firebase. Additionally, the new MongoDB extension enables real-time synchronization between Firebase and MongoDB data, ensuring data consistency across both platforms. By combining the power of Firebase Extensions, MongoDB Atlas, and a direct MongoDB connector, developers can create innovative and data-driven applications with greater efficiency and ease. Streamlining cloud migrations Developers are taking advantage of the latest models and tooling to build the next wave of gen AI applications. As they do so, they’re wrangling unprecedented volumes of structured and unstructured data, and, in doing so, are facing growing requirements for application scalability and performance. As such, businesses have uncovered a newfound imperative to modernize and take advantage of cloud-native solutions that offer the highest levels of scalability and performance, as well as interoperability with their favorite tools. MongoDB and Google Cloud have made it even easier to make the move to the cloud with Google Migration Center, a unified platform that streamlines the transition from on-premises servers to the Google Cloud environment, offering tools for discovery, assessment, and planning. Within Google Cloud Migration Center, users can now generate cost assessments and migration plans for the on-premises MongoDB instances directly in Google Cloud Migration Console, simplifying the transition to MongoDB Atlas on Google Cloud. Specifically, users can now use integrated MongoDB cluster assessment to gain in-depth visibility into MongoDB deployments, both Community and Enterprise editions, running on your existing infrastructure. Learn more on our blog . With the aim of making it easier and quicker to deploy to the cloud, MongoDB Atlas is now available as part of Cloud Foundation Fabric. Specifically, Atlas is available within Fabric FAST, which streamlines Google Cloud organization setup using a pre-defined enterprise-grade design and a Terraform reference implementation. By integrating MongoDB Atlas, enterprises can enhance production readiness for applications that require persistent data, offloading database management overhead while ensuring high availability, scalability, and security. This integration complements FAST’s infrastructure automation, enabling organizations to quickly deploy robust, data-driven applications within an accelerated Google Cloud environment, reducing the time needed to establish a fully functional, enterprise-level platform. Check out the Github repository . Optimizing analytics and archiving Despite many organizations’ focus on modernization and generative AI, analytics and data warehousing remain essential pillars of the enterprise workflow. Building on our existing integration with BigQuery, Google Cloud’s fully managed data warehouse, MongoDB Atlas now offers native JSON support for BigQuery, eliminating the need for complex data transformations. This enhancement significantly reduces operational costs, improves query performance, and enables businesses to analyze structured and unstructured data with greater flexibility and efficiency. The Dataflow template is now 'Generally Available' for MongoDB and Google Cloud customers. A key advantage of this pipeline lies in its ability to directly leverage BigQuery's powerful JSON functions on the MongoDB data loaded into BigQuery. This eliminates the need for a complex and time-consuming data transformation process as the JSON data within BigQuery can be queried and analyzed using standard BQML queries. Learn more about the new launch from our blog . In a big step forward for MongoDB, Atlas now supports Data Federation and Online Archive directly on Google Cloud. With these new additions, users can effortlessly manage and archive cold data and perform federated queries across Google Cloud storage, all from within their MongoDB Atlas console. This integration provides businesses with cost-effective data management and analysis capabilities. Upskilling for the AI era Earlier this year, MongoDB introduced a new Skill Badges program , offering focused credentials to help learners quickly master and validate their skills with MongoDB. These badges are an excellent way for developers, database administrators, and architects to demonstrate their dedication to skill development and continuous learning. In just 60-90 minutes, participants can learn new skills, finish a short assessment, and earn a shareable digital badge through Credly. They can then display this badge on platforms like LinkedIn to highlight their accomplishments and career development. At Google Next, attendees will have the opportunity to earn the RAG with MongoDB Skill Badge by using our self-paced Google Lab or by interacting directly with our experts in the Makerspace. This badge focuses on building Retrieval-Augmented Generation (RAG) applications, teaching participants how to integrate vector search and improve retrieval workflows to enhance apps powered by LLMs. Whether you prefer the guided support in the Makerspace or the flexibility of the self-paced lab, this hands-on experience will provide you with advanced skills that you can apply to your projects. The bottom line is that we’ve been busy! MongoDB’s deepening collaboration with Google Cloud continues to unlock new innovations across AI, application development, and cloud infrastructure. Stop by Booth #1240 at Google Cloud Next or join one of MongoDB’s featured sessions to explore these advancements and discover how MongoDB and Google Cloud are shaping the future of AI and data-driven applications. Head over to our MongoDB Atlas Learning Hub to boost your MongoDB skills.

April 9, 2025

Introducing New Navigation for MongoDB Atlas and Cloud Manager

MongoDB is excited to announce a major update to MongoDB Atlas and MongoDB Cloud Manager : a redesigned user experience that improves the workflow and navigation to access services and tools. This redesign ensures users can seamlessly navigate the Atlas and Cloud Manager platforms, intuitively accessing their most-used services and completing tasks more efficiently. Figure 1. Previous project-level homepage and primary side navigation in MongoDB Atlas. The Atlas platform has expanded exponentially since the last navigation redesign in 2020, with MongoDB introducing a plethora of new features and functionality, including Atlas Search and Vector Search , Atlas Charts , and Atlas Stream Processing . The latest navigation redesign has been architected from the outset to encompass these capabilities, addressing users' diverse needs—from monitoring deployments and managing billing to enhancing data visualization and enabling advanced search functionality—while delivering a streamlined, workflow-driven platform for users. Figure 2. Previous resource context (e.g., organization, project, cluster) for workflow tracking in MongoDB Atlas. Figure 3. Previous top navigation architecture in MongoDB Atlas. Starting two and a half years ago, MongoDB’s Design Strategy team began the redesign process by collecting customer feedback and engaging in dialogue. The team’s overall goal with the Atlas and Cloud Manager redesign was to create a holistic, seamlessly integrated platform that streamlined the developer experience. Figure 4. Redesigned homepage and primary side navigation at the project level in MongoDB Atlas. The redesigned navigation improves developers’ experience in the following ways: Workflow-focused architecture: The new architecture is clean and intuitive. It preserves developers’ “flow state” by guiding them through drill-down workflows. The new navigation prioritizes platform services, highlighting them based on the user’s workflow. This makes it easier for developers to focus on the most relevant tools for their current tasks, enabling them to work more efficiently and innovate faster. Consistent, familiar experience: The new navigation design provides a consistent experience across Atlas and Cloud Manager platforms. This makes it easier for developers to switch between the two interfaces. This consistent, intuitive interface enhances wayfinding and boosts overall productivity. What’s changing in MongoDB Atlas? The redesigned Atlas navigation introduces the following key updates: 1. Clearer resource context The updated top navigation bar, the resource navigator, ensures developers always know which resource (e.g., project, organization, cluster) they are working on. Switching between resources is now simpler, with improved context clarity as users navigate deeper into Atlas. For example, imagine switching between search indexes across different collections. Now, it can be done in a single click. The new workflow negates any need to backtrack to a project’s overview. Figure 5. Redesigned secondary side navigation at the project level in MongoDB Atlas, with an extended resource navigator in the top navigation bar. 2. Centralized utilities hub Essential utilities like Alerts, Billing, Help, and Identity Management (IAM) are consolidated in one location at the top-right corner. This ensures rapid access and saves time. Users can also access the product menu to find MongoDB University , Documentation , Community Forums , and Support . Figure 6. Redesigned utilities hub with an expanded product menu in MongoDB Atlas. 3. Simplified left navigation The side navigation is now organized into four categories: Database , Data , Services , and Security . These categories act as distinct containers, grouping Atlas’s capabilities to reflect the tasks a developer needs to perform within Atlas. This new structure makes navigating Atlas easier, helping developers find the right tools faster. Below is a breakdown of where features will be housed to make access to your essential tools even more straightforward: Database: Contains all core database capabilities. Includes cluster management and monitoring tools for browsing and querying, backups, and Online Archive . Data: Contains tools for working with data. Includes tools like visualization (Atlas Charts) to create and embed data visualizations, Atlas Search and Vector Search for powerful search capabilities and Data Federation for cross-source queries. Services: Contains features for event-driven data processing and automation. Includes capabilities such as Stream Processing for real-time data analysis, Triggers for automating database actions, and Migration for migrating existing deployments to Atlas. Security: Contains controls for data access and protection. Includes capabilities like project settings, Identity & Access Management (IAM), auditing, and advanced security. On the organizational level, the new architecture for the side navigation will be organized into two categories: identity and access , and billing . Figure 7. Redesigned primary side navigation at the organization level in MongoDB Atlas. What’s changing in Cloud Manager? Although Cloud Manager will function similarly to Atlas, we did make several changes to refine the Cloud Manager experience: 1. Left navigation On the organizational level, the new architecture for the side navigation will be organized into three categories: Identity and access: Add, delete, and manage users, teams, and API Keys within a specific project. Billing: View, track, and manage your charges while using Cloud Manager. Management: Set up Kubernetes , and manage additional administrative functions. Figure 8. Redesigned primary side navigation at the organization level in Cloud Manager. On a project level, the side navigation will be organized into two categories: Database: Manage Processes, Servers, Agents, Security, and Continuous Backup for your deployments. Admin: Monitor Pings, MongoDB Process Arguments, Deleted Hosts, Profiler Request History, and Raw Automation Config. Figure 9. Redesigned primary side navigation at the project level in Cloud Manager. 2. Resource navigator Cloud Manager will have the same resource navigator tool in the top navigation bar as Atlas. This provides clear visibility of the resource users are working on in Cloud Manager, whether it’s a project or an organization. 3. Centralized utilities hub Cloud Manager will also feature the updated utility hub mentioned in the changes coming to Atlas. This hub allows users to access the same essential utilities and product menu to discover other MongoDB offerings in one place. Rollout timeline To ensure a smooth transition, MongoDB will be rolling out the new navigation experience in phases. The Atlas update is currently going live, and the Cloud Manager update will begin the week of May 12, 2025 . Note that the Atlas experience will take 6 to 8 weeks to be available to all Atlas organizations. All organizations will experience the new navigation by June 2025 . Atlas and Cloud Manager users can submit feedback to share their thoughts on the new navigation experience. Explore MongoDB’s updated documentation for more details on the latest changes to the navigation. Try the new navigation today through your MongoDB Atlas or Cloud Manager portal.

April 8, 2025