Intercom Doubles R&D Throughput with AI-First Engineering Approach
In a remarkable transformation, Intercom has doubled its research and development (R&D) throughput in just nine months. The company achieved this feat by embracing an AI-first engineering approach, led by senior principal engineer Brian Scanlan. This shift enabled 100% of engineers to ship code via Claude Code.
The key statistics include:
- A 100% increase in merged PRs per R&D employee
- A doubling of R&D throughput in nine months
To measure the success of their AI adoption, Intercom built a telemetry infrastructure using Honeycomb. This system tracks skill usage across hundreds of engineers and provides personalized insights into session data collection.
The company's skills repository is designed to enforce engineering standards automatically through hooks. Scanlan also discussed how Intercom prepares its product for an agent-first world with CLIs, MCPs, and ephemeral APIs.
