ani.tech cuts 26-hour workflows by 75% with AI

THE CHALLENGE

Overcoming data layer rigidity to feed autonomous AI

In the fast-moving, security-sensitive world of financial services, legacy data constraints can often stand in the way of product innovation. Against this backdrop, Samantha McBride founded ani.tech, an AI-first fintech company helping investment and wealth management firms to automate complex advisory and portfolio management workflows.

Originally launched, as McBride put it, as a ‘classic software-as-a-service (SaaS) platform,’ the startup was built on a traditional Django and MySQL framework. But convincing traditional financial services clients to abandon their Excel spreadsheets for software wasn’t moving at pace. Sensing a massive technological shift, and realizing an opportunity to completely redefine the business, McBride and the team secured early beta access to generative AI models. They envisioned a future powered by autonomous AI workforces capable of automating complex financial workflows. However, pivoting the business from SaaS platform to agentic AI revealed limitations in the traditional relational database foundation it had previously relied on.

The engineering team faced constant friction; every new product iteration triggered time-consuming schema migrations and sluggish deployment cycles that choked momentum. Forcing sophisticated, dynamic AI capabilities onto a rigid technology stack proved highly inefficient. Relational data layers simply could not deliver the fluid flexibility required to feed autonomous AI agents.

“We realized early on that just bolting these technologies on to the legacy tech stack was less than ideal,” said McBride. “We want every element of our tech stack to be AI native, so we completely rebuilt it from the ground up. And MongoDB was one of the big changes we made.”

OUR SOLUTION

Building an AI-native operational platform in the cloud

To eliminate legacy database constraints and build an AI-native infrastructure, ani.tech strategically chose MongoDB Atlas as its primary operational data platform, deploying it on Microsoft Azure to gain maximum flexibility. This highly scalable architecture now underpins the company’s advanced Gen AI workflows, serving as the foundation for their autonomous workers, orchestration layer, and internal guardrails.

To support this rich ecosystem, ani.tech split its environment across a core application database and a data warehouse, effortlessly managing approximately 22.6 million documents across 202 collections. Within this fluid data layer, autonomous agents use MongoDB to store real-time chat histories and evolving workflow states. At the same time, specialized tool-calling agents query the database directly to surface complex financial insights and risk exposure data in real time. MongoDB Atlas is also used to underpin heavy time-series workloads, continuously ingesting long-range financial datasets.

Managing such massive, concurrent data pipelines requires highly intelligent infrastructure. ani.tech runs its production environment on a scalable cluster architecture that can automatically expand to handle heavy ETL (Extract, Transform, Load) and analytical workloads without compromising performance.

By using MongoDB as a flexible data layer on top of existing CRM and operational systems, ani.tech allows financial institutions to adopt AI capabilities without overhauling the strict, decades-old infrastructure that such services often rely on. The platform bridges fragmented compliance and financial data across legacy systems, restructuring it in real time so that agents can integrate with it.

“The great thing about utilizing MongoDB and agentic AI is that you don't have to worry about how your data is stored underneath,” said McBride. “We can offer a flexible layer that transforms clients’ data into the structure and data sets our agents need to complete tasks.”

OUTCOME

Unlocking rapid iteration and massive operational efficiencies 

By re-architecting on MongoDB Atlas, ani.tech has significantly improved both development speed and client onboarding. The company now relies heavily on AI agents as part of its development process—building them for both internal use and external clients—and its “fail fast, iterate quickly” philosophy is, today, fully enabled by MongoDB.

Historically, development cycles were slowed by the need to design and migrate database schemas, build APIs, and coordinate handoffs to front-end teams. Now, with AI agents also embedded into the deployment process, the team can ship updates much faster. “We see it as trying to do things on an hourly basis, as opposed to days, weeks, months,” said McBride. “MongoDB just gets out of the way. You don’t have to worry about database migrations or schema design—you can just prototype, ship quickly, and iterate from there.”

When ani.tech deployed its AI agents for its pilot production customer, Octopus Money, it achieved massive operational efficiencies. Its automated agents successfully absorbed 75% of an end-to-end financial-advice workflow, including report and portfolio creation, that previously required 26 hours of manual processing. “So, we think that 75% benchmark is a good target for deciding which workflows are AI-friendly,” said McBride.

Beyond sheer speed, partnering with MongoDB has enabled ani.tech to simplify technical due diligence for its financial services clients, for whom security is non-negotiable. “The great thing about MongoDB is that it's enterprise-grade, it's very well-known and credible in the industry, and all of the security features we need are out of the box,” said McBride. “It provides an easy checkbox.”

Backed by this robust, flexible foundation, ani.tech is currently onboarding a further five major financial services clients, a pipeline that will trigger a fivefold increase in the company's revenue and data requirements.

Ultimately, ani.tech’s partnership with MongoDB has redefined how the company scales its operations, making speed and iteration key enablers. “The flexible nature of MongoDB means that it's very AI agent friendly,” concluded McBride. “And that gives us flexibility as we grow as well.”