How a personal AI productivity hack became Slack's internal tool

by Frances Coronel

Frances Coronel, a senior software engineer at Slack, presents the catalog in this GitHub Universe 2026 session, demonstrating how it works and sharing six months of adoption metrics.

From personal setup to shared infrastructure

The effort began with Coronel building a personal AI setup to work faster and sharing it in a public Slack channel. Other engineers started forking it without being asked. That organic interest grew into an internal AI tool catalog intended to help teams discover and share tooling rather than maintain separate pockets of knowledge.

A catalog built on GitHub

The session introduces Slack’s DevX PI catalog as a single source of truth for:

Coronel demonstrates the catalog and covers its architecture and growth, including how GitHub Projects and Issues support an internal tool catalog. The session also explores experimentation and automation with AI, along with accessing the catalog through Slack.

Measuring adoption and encouraging sharing

An adoption dashboard tracks real usage data across teams. Coronel shares six months of metrics and discusses measuring adoption through telemetry rather than surveys.

Alongside the tooling, the session addresses AI knowledge silos and the cultural conditions that help useful setups spread organically. It closes with recommendations for internal AI tools, connecting discovery, sharing, and usage measurement to the broader work of supporting AI adoption across engineering teams.

Session chapters