AI Coding at Scale: Working With Large Codebases and Monorepos — Tips and Tricks

AI coding agents are genuinely powerful, and on a small project they feel almost effortless. Point them at a big codebase — a monorepo with a dozen packages, years of history, several teams — and you start to notice the rough edges: context fills up fast, conventions can slip, and the agent isn't always sure which part of the code to focus on. None of it is a dealbreaker; it just takes a different approach.
This meetup is about that approach: the setup and the habits that keep AI coding agents useful once the codebase gets big.

Evgeny Potapov, engineering manager with over 20 years of experience and cofounder of US-based startup ApexData, will share hands-on patterns from real work with Claude Code (and a bit of Codex):

— How it's even possible: a look under the hood at how Claude Code and Codex operate on a codebase far larger than their context window — searching and reading on demand, and using sub-agents so the main context stays clean
— Keeping the agent focused and well-fed: managing context, and how giving it the right related codebases and services actually makes it more capable
— Pulling several projects into context: the backend, the app that produces the data, and product docs together, then a PoC of a feature that uses data from other apps in the project
— Practical cases: onboarding into an unfamiliar codebase, tracing how things connect across the repo, and making changes that cut across many packages at once

If you're a developer, tech lead, or engineering manager already using AI coding tools and curious how far they can go on bigger projects, come for concrete patterns and a chance to compare notes with others working at the same scale.

🕒 Schedule:
18:30 🥂 Networking, Pizza & Drinks
19:00 👋 Welcome Words
19:15–20:00 🎤 Evgeny Potapov: AI coding at scale: working with large codebases and monorepos — tips and tricks