Wes McKinney's Agentic Engineering: 3-Person Team Merges Hundreds of PRs Weekly
josh_wills · x · 2026-08-12
In a detailed blog post, Pandas creator Wes McKinney shared insights into the agentic engineering process and culture at his startup, Kenn Software.
- Core Philosophy: He strongly criticized fully autonomous "loop engineering," calling bullshit on the idea that you can step away from the keyboard and get quality output from agents looping on each other. He differentiates his team's approach of "planning, architecture, and caring about the output" from mere "vibe coding."
- Massive Output: With a team of just three people, they merge hundreds of pull requests per week while maintaining an empirically low bug rate across millions of lines of production code.
- Token Consumption: McKinney is a heavy AI user, reporting $56,836 in API token costs over the last 30 days alone (subsidized by coding agent subscriptions).
The post emphasizes that human taste, architectural control, and genuine care for the output remain the core guarantees of quality in AI-assisted development.
More from coding & agent
- Ex-Cognition & OpenAI Members Launch Hone for Org-Level AI Agents — marvinvonhagen · 2026-08-13
- Trust AI coding agents to ship production code without human review? — meghna_rana · 2026-08-13
- Handing Over an NFL Fantasy Football Team to AI for a Full Season — Prestigious-Dig2263 · 2026-08-13
- Kimi K3 Hits GitHub Copilot: Developers Build Games from a Single Prompt — film_girl · 2026-08-13
- Meta Launches Muse Code AI Agent with Pay-As-You-Go Pricing to Rival OpenAI — Beth_Kindig · 2026-08-13
- DeepMind Researcher Tests Claude Code: AI Agents Are Solving Real Problems — mpshanahan · 2026-08-13