🎙️ This week on How I AI: How Intercom 2x’d their engineering velocity with Claude Code

04/20/26
Source: Lenny’s Newsletter by Lenny Rachitsky. Read the original article

TL;DR Summary of How Intercom Doubled Engineering Throughput Using Claude Code AI

Intercom doubled engineering velocity in nine months by fully integrating Claude Code AI with custom skills and deep telemetry. They treat their engineering org like a product by instrumenting every interaction to gain actionable insights and enforce quality. A culture of permission and accountability empowers engineers to experiment and accelerate workflows, while fixing fundamentals ensures AI amplifies strengths without sacrificing quality.

Optimixed’s Overview: Leveraging AI and Culture to Transform Engineering Productivity

Engineering Throughput Multiplied by AI Integration

Intercom achieved a remarkable 2x increase in merged pull requests per R&D employee within nine months by going all-in on AI-powered tooling with Claude Code. This success stemmed from combining mature development practices with innovative AI adoption strategies:

  • Custom Skills & Guardrails: Creating AI hooks that enforce quality at the point of creation, such as context-rich pull request descriptions, ensures shipping high-quality code faster.
  • Instrumentation & Telemetry: Tracking AI usage via Honeycomb and storing anonymized sessions in S3 enables visibility into what works and where improvements are needed, applying product thinking internally.
  • Culture of Permission: Leadership encourages experimentation by granting engineers freedom to connect AI tools with infrastructure, shifting activation energy from technical to cultural barriers.

Fix Fundamentals Before Scaling AI

AI acts as a force multiplier that amplifies both strengths and weaknesses. Intercom emphasized the necessity of having:

  • Mature CI/CD pipelines
  • Comprehensive test coverage
  • A high-trust culture

Without these, AI would only accelerate existing problems. By addressing foundational issues first, they ensured rapid scaling of AI benefits without compromising code quality.

Driving Business Impact and Future-Ready Workflows

Beyond engineering velocity, AI adoption enabled faster resolution of internal projects and technical debt, leading to improved developer experience and higher code quality metrics supported by partnerships with academic researchers. Intercom also pioneered an autonomous CLI agent capable of complex workflows like email verification and installation without human intervention.

The vision is clear: all work will become agent-first, with AI agents responding to alarms, meetings, and customer queries as a baseline expectation. The key bottleneck is not technology but organizational willingness to rethink workflows and empower experimentation.