Conferences, meetups and sessions run inside engineering teams, on agent architecture, evaluation, and what production asks of both.
LangChain Interrupt · San Francisco · 2026
How an agent platform moved from a hierarchy of agents to a single flat one, traded a large tool catalogue for a handful of general tools, and let the model write SQL against the schema directly. Presented with Akash Ashok on Rippling's agent platform.
LangChain Meetup · Dubai · September 2025
Software engineering practice for agent systems, and what has to change when an agent stops being a demo and starts being something on call.
We help backend and fullstack engineers become AI engineers. They have spent years building systems that behave the same way every time. AI does not work like that, and it takes some getting used to. So we teach it by building. Your team works through one project with us, picks up the ideas as they go, and keeps everything at the end.
The big shift. A function you can unit test becomes a system that is right most of the time. That changes how you design, test and ship.
How the models actually behave. Context, tokens, temperature, structured output, and what an embedding really is. Enough theory to debug things, no more.
Chunking, indexing and hybrid search, built by hand. Where simple semantic search falls over, and how you would notice.
Tools, planning and control flow. When you need an agent, when one plain call will do, and why the neat diagram rarely survives contact with users.
How you ship and watch a system like this. Build a golden set, score what matters, run it in CI, and trace any answer back to the call behind it.
What a feature costs per request, and at ten times the traffic. Latency budgets, caching, model routing, and how to spot when either slips.
We agree the length and the mix before we start. Teams already shipping AI spend more of it on evals, cost and monitoring. Teams starting out spend more on retrieval and on the change in thinking. On-site or remote, whichever suits you.
Tell us the audience, or the team size and where they are starting from, and we will come back with a shape for the session.