How to run coding agents reliably in large codebases
Coding agents rarely fail because of the model alone.
In large codebases, the harder problems are operational: context drift across long-running tasks, retrieval across repositories, brittle tool chains, and review bottlenecks.
What you'll learn
- Why coding agents struggle in large codebases
- How teams are improving codebase context and retrieval
- Patterns for more reliable long-running agent workflows
- Ways to scale human oversight without slowing engineering teams down
Speakers
Stephanie Jarmak Agent Advocate at Sourcegraph