Kristen Min Chan

Tableau permission auditIn production

A monthly audit that was not realistic by hand, automated with Python, REST and GraphQL.

Under 1 hour of my time per monthly audit

In productionA colleague runs it every month from the SOP I wrote.

Power BI usage optimizationIn production

Monitoring that remembers past 14 days, an alert in real time, then the fix.

−51% in one workspace 13% of total enterprise capacity freed

In productionColleagues maintain the report from the SOP I wrote.

streamwatchIn production

Finds each new recording, waits until it is really finished, and sends one ready-to-publish email.

~52 actions by hand per item, before automation

In productionWeekly since August 2026, for one organization.

New data platform, enabled with AICompleted

200+ colleagues moved to a new data platform with built-in AI, with AI bridging the SQL differences.

200+ colleagues champions now onboard their own teams

Usage Surge Alert AgentDemo

The capacity problem rebuilt as an AI workflow. Personal data masked before any model call.

Every send approved by a person first

kdpgenOpen source

Generative AI for Kindle Direct Publishing (KDP): an open-source pipeline that turns one JSON spec into print-ready book files. AI generates the art; a person picks every page.

Proofed in print output approved from physical proof copies

Automating MinOngoing

Learning in public: AI automation write-ups with real numbers, plus the brand system and the self-built Claude skills that enforce it.

Build and learn in public