Ramp runs AI agents across its entire engineering lifecycle: writing code, reviewing it, watching production, and root-causing incidents. Boris sat down with Austin Ray and Rahul Sengottuvelu of Ramp to talk about how they got there. Building for the models that are coming rather than the ones that exist, giving every engineer uncapped access to intelligence, and the guardrails that make it work. They compare notes on Claude Code setups, loops versus dynamic workflows, and what Claude Fable 5 unlocked.
Claude Code: anthropic.com/product/claude-code Claude Cowork: anthropic.com/product/claude-cowork Office Hours LP: claude.com/office-hours
Chapters 0:00 "Fix all our import cycles" 0:32 Stress-testing Fable on Ramp's Python modules 1:33 Fable and dynamic workflows cut CI time 66% 3:36 Loops vs. dynamic workflows for long-horizon tasks 5:15 Claude Code setups: vanilla vs. background-heavy 6:49 AI agents across the engineering lifecycle 7:23 Building for future models, not today's 9:11 AI agent guardrails and least privilege 12:00 Cost controls and AI code review 13:08 Ramp's culture of experimentation 13:52 Glass and Inspect: Ramp's AI coworkers 16:05 On-call assistant: an AI SRE on Claude Code 17:13 More agent sessions from automations than humans 18:44 No token budgets for engineers 20:48 Advice for CTOs adopting AI agents
