How Amplitude built an internal AI tool that the whole company’s obsessed with (and how you can too) | Wade Chambers
How I AI
2025/08/11
How Amplitude built an internal AI tool that the whole company’s obsessed with (and how you can too) | Wade Chambers
How Amplitude built an internal AI tool that the whole company’s obsessed with (and how you can too) | Wade Chambers
How I AI
2025/08/11
In this episode, Claire Vo sits down with Wade Chambers, Chief Engineering Officer at Amplitude, to explore how his team rapidly developed Moda—an internal AI tool that revolutionized access to enterprise data and streamlined product development workflows. The conversation dives into the technical and cultural strategies behind Moda’s creation, as well as its impact on cross-functional collaboration and decision-making.
Wade Chambers explains how Amplitude built Moda in just a few weeks using spare engineering time and external APIs like Glean. The tool allows employees to access internal data across systems and generate PRDs, accelerating product development. A key factor in its rapid adoption was a social engineering approach that leveraged visibility and peer influence. Moda integrates with Slack and pulls from sources like Confluence and Jira, while maintaining strict data privacy. Product managers use it to analyze customer feedback and identify trends. The tool also enables role-swapping across teams, fostering empathy and collaboration. AI-generated PRDs are reviewed by humans, ensuring quality and alignment. Moda has compressed workflows that previously took weeks into a single meeting, significantly increasing development speed. Finally, Wade touches on how AI is helping engineers manage technical debt and improve efficiency.
04:30
04:30
Wade believes building internally is worth the effort if not over-engineered
04:56
04:56
Moda was built in 3-4 weeks and helps unlock internal data for PRDs and other artifacts
07:19
07:19
The tool's public visibility, ability to accelerate thinking, and the social proof of colleagues' use contributed to its rapid adoption.
09:18
09:18
Moda can search internal and public knowledge sources, cites sources, and can be accessed via Slack or a proprietary interface.
10:58
10:58
Multiple company departments use the AI product for various purposes.
15:54
15:54
AI startups need more than speed; security, compliance, and transparency are crucial
17:49
17:49
Moda's proprietary interface allows product managers to create PRDs and do deep research.
23:24
23:24
Moda expands a prompt into a full PRD with detailed requirements and prototypes
27:30
27:30
AI-generated documents require human review to ensure quality and alignment.
29:09
29:09
Using Moda, tasks that previously took weeks can now be completed in a single meeting, increasing overall team velocity.
32:37
32:37
Team members swap roles in a demo, with designers acting as engineers, engineers as product managers, and product managers as designers.
34:38
34:38
Wade built Moda, an internal AI tool, in a few weeks with minimal resources.
36:03
36:03
AI can help deal with tech debt more effectively and reduce engineers' toil.
