Intercom Revolutionizes Its Engineering with Claude Code in 9 Months

⚡

Key Takeaways

1Intercom doubled its R&D productivity in nine months thanks to Claude Code, involving 100% of its engineers.

2A telemetry infrastructure has been established to track AI adoption and code quality at Intercom.

3The company has created a skills repository to automate engineering standards and prepare its product for an agent-focused future.

💡Why it matters — This transformation demonstrates how the integration of AI can radically improve efficiency and quality in the tech industry.

Intercom has recently made a significant breakthrough in its engineering process by doubling its development speed in just nine months. This transformation was made possible through the integration of Claude Code, an artificial intelligence tool that enabled 100% of the company's engineers, designers, PMs, and TPMs to participate in code shipping.

Brian Scanlan, a principal engineer at Intercom, led this transition towards AI-driven engineering. In nine months, the company doubled the number of PRs merged per R&D employee while maintaining code quality. This achievement was realized through the use of Claude Code, which facilitated widespread AI adoption among Intercom teams.

An Advanced Telemetry Infrastructure

To measure the impact of AI on its operations, Intercom has established a sophisticated telemetry infrastructure. This system allows for tracking AI adoption and assessing the quality of code produced by hundreds of engineers. This approach has enabled the company to maintain high standards while increasing productivity.

An Innovative Skills Repository

Intercom has also developed a skills repository equipped with hooks that automatically apply engineering standards. This innovation has facilitated rapid AI adoption and prepared the company for a future where agents will play a central role, thanks to the use of ephemeral CLIs, MCPs, and APIs.

Towards a Zero Backlog

One of Intercom's ambitious goals is to achieve a zero backlog, a concept now feasible due to the increased efficiency brought by AI. This shift has profound implications for the company's engineering culture, allowing for more agile and responsive project management.

Rapid Adoption and Accountability Framework

The success of this transformation also relies on a well-defined framework of permission and accountability, which has facilitated the rapid adoption of AI within the company. This framework has allowed for treating AI expenditures as a strategic investment rather than just a cost.

Tools and References

Among the tools used in this transformation are Claude Code, Cursor, Honeycomb, Snowflake, Fin AI, and Vercel. Intercom's GitHub repository and the Google API Go client have also been key resources in this process.

To learn more about the key players in this transformation, you can follow Brian Scanlan on X and LinkedIn, as well as Claire Vo on ChatPRD, her website, LinkedIn, and X.