Rocky Mountain Ruby 2026
High-Octane Onboarding: Ramping into a Rails Codebase with AI Agents
Joining a new Rails codebase used to mean weeks of reading code and slowly building a mental model before you could really contribute. For me, ramping up at Zipline this year, that math changed
This is a field report from my first 90 days as a Senior Rails engineer, and the AI-assisted practices that let me actually ship value while I was still figuring out where everything lived. We'll cover the artifacts that prime both humans and agents for context (think personal and team READMEs you can hand to an LLM), MCP integrations that make Jira and Confluence queryable, and using agents for codebase archaeology when you're three layers deep in a service object.
You'll leave with a playbook you can actually use, plus the honest gotchas, because the agents will lie to you with full confidence.
This is a field report from my first 90 days as a Senior Rails engineer, and the AI-assisted practices that let me actually ship value while I was still figuring out where everything lived. We'll cover the artifacts that prime both humans and agents for context (think personal and team READMEs you can hand to an LLM), MCP integrations that make Jira and Confluence queryable, and using agents for codebase archaeology when you're three layers deep in a service object.
You'll leave with a playbook you can actually use, plus the honest gotchas, because the agents will lie to you with full confidence.