On May 7th, Keith (Quality Engineer, Percona for MongoDB) and I had the super cool opportunity to head over to MongoDB.local London! The event was amazing and packed with insights about where the database ecosystem is heading.

If there was one massive takeaway from the day, it was this: We are officially in the Era of AI and “Agentic” Systems. During the event, the message was clear: we are shifting from basic LLMs (that just answer a prompt and forget it) to autonomous AI Agents that follow a continuous loop of Perception → Planning → Action. MongoDB’s President and CEO, CJ Desai, repeated a powerful phrase:

While AI models change rapidly, the Data Layer is the constant.

Here is a look at our day, what we learned, and the fun we had along the way!

We got a great tip before the event: arrive early to get a head start on the gamified learning! MongoDB had a super nice setup where you could take tests on Credly to earn knowledge badges.

We jumped right in. I got a MongoDB Overview badge, but Keith was on a mission. He completed three different tests (including MongoDB for Developers) and unlocked some cool swag: a really cute, high-quality bag! It was a brilliant way to get attendees engaged right from the morning.

Here are more badges in case you want to get yours!

We spent a lot of our time in the main room for the General Session, and the announcements were packed with impressive numbers and tech:

• MongoDB 8.3 is Fast: Osmar Olivo (Senior Director, Database Product Management) shared that the new version brings up to 35% more write throughput, 45% more read throughput, and 15% more for ACID transactions.
• The Scale is Real: We learned that Stripe uses MongoDB to process over $1 trillion in payments volume every year (maintaining 5 nines of availability!). Osmar framed this perfectly: that is 1.5% of the global GDP running through MongoDB.
• LangGraph.js Store Integration: This was a big one for developers. MongoDB is positioning itself as the “memory hard drive” for AI agents. By supporting JavaScript and TypeScript, they are making it super easy for companies to use their existing web developers to build complex AI workflows.
• Hugging Face Partnership: They are scaling with MongoDB Atlas to support over 3 million models, officially tying themselves to the “GitHub for AI.”

Feel free to explore the recorded sessions for more: MongoDB.local London 2026

Ulku Rowe (CIO, Commercial Business at Lloyds Banking Group) talked about this being the “Decade of AI.” Lloyds is actively upskilling their current engineers through an internal “AI Academy” built in partnership with Cambridge University! She emphasized that as they build out this infrastructure, partnerships are absolutely critical to their success.

We also heard from Alex Holt from ElevenLabs, a company focused on producing the absolute best, human-sounding voice AI. Their scale is wild: they have 40 million agents running and hit $500 million in Annual Recurring Revenue in just 3 years! Alex mentioned that because many enterprises don’t know how to build agents yet, ElevenLabs uses “forward deployed engineers” to sit directly with customers to build, deploy, and prove the ROI of their voice agents.

Later in the day, we attended a hands-on workshop: Designing Memory Systems for AI Agents, hands-on workshop about how AI agents can remember information and use it later to give better responses. We used Python and MongoDB Atlas to build memory into an AI agent and learned how to store, search, update, and manage that memory. The setup was good, everything was prepared in advance so we could focus on executing the commands and truly understanding the concepts. At the end, we answered some questions and earned another badge!

However, the workshop ran until 1:00 PM. One of our friends had warned us to “go for food fast,” but we were too focused on the workshop! By the time we made it to the lunch area, all the main food was completely sold out.

How did we survive? Chips, candies, and a lot of beverages. Between the sodas, coffee, and tea, we kept our energy, but it was definitely a funny learning experience for next time! I can imagine Keith arriving home for dinner!!

We spent our afternoon speaking with sponsors and even got to participate in a quick podcast focusing on AI and how Atlas is being used as a strong platform for these projects!

The person being interviewed was Bikram Das, who is Chief Data Architect at Tata Consulting Services, and we had a great chat with him. TCS and MongoDB have partnered on a super impressive real-time payment and fraud detection platform. They use autonomous AI agents to instantly assess risk, investigate anomalies, and route safe transactions to networks like Visa and SWIFT without any downtime.

We also talked with IBM folks; they showed us their “plug-and-play” enterprise AI foundation. They are focused on letting large companies safely deploy AI agents without having to completely rip out and rebuild their current data infrastructure.

We also stopped by the Accenture booth! They are actively working on integrating AI directly into their platforms so they can offer smarter, more advanced solutions to their customers.

To cap off a great day, MongoDB had one last treat. If you took less than 3 minutes to fill out the end-of-event survey, they handed you a super nice pair of socks. (We love community ideas like this!).

Overall, MongoDB.local London was a great experience. It was a nice space to learn, connect, have hands-on experience, and see exactly how the database world is evolving to meet the Agentic AI era head-on.

See you at the next event!

∎