Director of Product at Google Labs leading AI Developer Assistants (AIDA). Previously VP of Product at Vercel, Senior Director at GitHub, and product leader at Heroku. Over a decade of experience building developer tools and platforms that developers love.
Champion of Proactive, Ambient AI Agents
Kath Korevec has emerged as a leading voice advocating for the shift from reactive chat-based AI to proactive, ambient agents that anticipate developer needs. Her philosophy: “The patterns of how we use an IDE right now might not exist at all next year.”
Current Work: Google Labs AIDA
At Google Labs, Kath leads the AI Developer Assistants team building next-generation autonomous coding tools:
- Jules - Autonomous coding agent powered by Gemini 2.5 Pro, handling async tasks from bug fixes to performance optimization. Now available via jules.google with CLI tools and API access.
- Stitch - AI-powered app design tool at stitchai.org for rapidly creating UI from text or image prompts.
Her team focuses on proactive agents that observe context, take timely action, personalize to individual developers, and integrate seamlessly into existing workflows.
Background
Career trajectory:
- Google Labs - Director of Product, AIDA team
- Vercel - VP of Product & Head of Product, led Next.js ecosystem strategy
- GitHub - Senior Director, led Community product team
- Heroku - Early developer tools experience
From Alaska (north of Anchorage), now based in San Francisco. Passionate about making world-class software development accessible to anyone, anywhere.
Philosophy: The Mental Load Problem
Kath frames AI agents through a relatable metaphor: just as someone needs to remind their husband to do the dishes, developers face significant mental load managing repetitive tasks. Her vision centers on four pillars:
- Observation - Understanding development context continuously
- Timely Action - Acting proactively at the right moment
- Personalization - Learning individual developer patterns
- Seamless Integration - Embedding into workflows without friction
Key insight: When compute becomes unlimited, agents should handle mechanical work so humans can focus on creativity and architecture.
Notable Projects & Content
Jules in Production:
- Approaching 2M developers using autonomous coding features
- Real-world case study: Building a 6-foot animatronic Halloween head with Jules handling firmware and stepper motor programming
- Developer freed to focus on creative design while Jules managed tedious implementation
Writing & Thought Leadership:
- Kathy PM (Substack): kathykorevec.substack.com - DevEx strategies and lessons from building developer tools
- GitHub README Stories: Featured profile on building developer solutions
- Active on X/Twitter: @simpsoka
Conference Appearance
Event: AI Engineering Code Summit 2025 Date: November 20, 2025 Time: 1:45 PM Session: “Proactive Agents: Moving Beyond Chat-Based AI”
Presented vision for autonomous agents that reduce cognitive load, demonstrated Jules’ self-directed task generation, and challenged developers to question traditional IDE patterns as AI reshapes development workflows.
Key Resources
- Jules: https://jules.google - Autonomous AI coding agent
- Jules Blog: https://blog.google/technology/google-labs/jules/ - Official announcements and product updates
- Stitch: https://stitchai.org/ - AI app design tool
- Substack: https://kathykorevec.substack.com/ - Developer experience insights