MongoDB at
AI Engineer World’s Fair 2025

In-booth sessions:

GraphRAG: Integrating LLMs with Knowledge Graphs (Thibaut Gourdel, Senior Technical Product Marketing Manager)

Smarter Together: Designing Multi-Agent Systems with Shared, Evolving Memory (Mikiko Bazeley, Staff Developer Advocate)

RAG in 2025: State of the Art and the Road Forward

Speaker: Tengyu Ma, Chief AI Scientist at MongoDB

The talk will have three parts:

1.Roadmap debate: RAG vs. finetuning vs. long-context
2.RAG today: benefits, challenges, and current solutions
3.RAG tomorrow: AI models do more work

The State of AI-Powered Search and Retrieval

Speaker: Frank Liu, Staff Product Manager at MongoDB

In this talk, we examine the state-of-the-art in AI-powered search and retrieval. We detail techniques for enhancing performance beyond base embedding models, including hybrid search, reranking strategies, query decomposition and document enrichment, the use of domain-specific and fine-tuned embeddings, custom data processing pipelines (ETL), and contextualized chunking methods.

Architecting Agent Memory: Principles, Patterns, and Best Practices

Speaker: Richmond Alake, Staff Developer Advocate, AI/ML at MongoDB

In the rapidly evolving landscape of agentic systems, memory management has emerged as a key pillar for building intelligent, context-aware AI Agents. Inspired by the complexity of human memory systems—such as episodic, working, semantic, and procedural memory—this talk unpacks how AI agents can achieve believability, reliability, and capability by retaining and reasoning over past experiences.

We’ll begin by establishing a conceptual framework based on real-world implementations from memory management libraries and system architectures:

• Memory Components representing various structured memory types (e.g., conversation, workflow, episodic, persona)
• Memory Modes reflecting operational strategies for short-term, long-term, and dynamic memory handling

Next, the talk transitions to practical implementation patterns critical for effective memory lifecycle management:

• Maintaining rich conversation history and contextual awareness
• Persistence strategies leveraging vector databases and hybrid search
• Memory augmentation using embeddings, relevance scoring, and semantic retrieval
• Production-ready practices for scaling memory in multi-agent ecosystems

We’ll also examine advanced memory strategies within agentic systems:

• Memory cascading and selective deletion
• Integration of tool use and persona memory
• Optimizing performance around memory retrieval and LLM context window limits

Whether you're developing autonomous agents, chatbots, or complex workflow orchestration systems, this talk offers knowledge and tactical insights for building AI that can remember, adapt, and improve over time.

This session is ideal for:

• AI engineers and agent framework developers *Architects designing Agentic RAG or multi-agent systems
• Practitioners building contextual, personalized AI experiences

By the end of the session, you’ll understand how to leverage memory as a strategic asset in agentic design—and walk away ready to build agents that not only act and reason but also remember.

Building Multimodal AI Agents (From Scratch)

Speaker: Apoorva Joshi, Sr. AI Developer Advocate at MongoDB

In this hands-on workshop, you will build a multimodal AI agent capable of processing mixed-media content—from analyzing charts and diagrams to extracting insights from documents with embedded visuals. Using MongoDB as a vector database and memory store, and Google's Gemini for multimodal reasoning, you will gain hands-on experience with multimodal data processing pipelines and agent orchestration patterns by implementing core components directly, using good ol' Python.

Prerequisites: Basic Python experience.

Materials: Complete GitHub repository with code and learning materials will be provided.

Smarter Together: Designing Multi-Agent Systems with Shared, Evolving Memory

Speaker: Mikiko Bazeley, Staff Developer Advocate at MongoDB

In today’s most advanced AI systems, intelligence is no longer confined to a single model or agent—it emerges from coordination. But coordination requires memory: short-term, long-term, and shared. In this talk, we’ll break down how agent systems can store, retrieve, and evolve shared memory to become smarter over time. You'll learn what it takes to architect these continuously learning systems, how to track and improve memory quality, and why robust, flexible infrastructure is the foundation of it all. Stick around to see how this works in practice—live.

GraphRAG: Integrating LLMs with Knowledge Graphs

Speaker: Thibaut Gourdel, Senior Technical Product Marketing Manager at MongoDB

While traditional RAG is effective, it can struggle with complex relationships and reasoning across large knowledge bases. GraphRAG, an advanced variant, addresses these challenges by leveraging knowledge graphs to enable deeper understanding and improved response accuracy. Learn how LLMs extract key entities and relationships from your data to construct a graph structure, and how the system uses graph traversal to find related entities and enrich prompts. Stay for a live demo showcasing these concepts in action.

Vectors as Datatype; How Modern Databases have Evolved in the AI Age

Speaker: Henry Weller, Senior Product Manager at MongoDB

Learn new techniques of indexing and querying with vectorSearch, including querying multiple vectors in one query, indexing and searching images, and indexing slices of embeddings from a single vector in your collection.