How software engineering is changing: a tale of two eras

Founding engineer Neil Kakkar shows that at Tano an AI review agent replaced human code review, raising his output elevenfold while shared knowledge and focus suffered.

Neil Kakkar, a founding engineer at Tano, an influencer marketing agency that builds AI agents to negotiate with content creators, won The Pragmatic Engineer's essay prize. He compares his output at the financial data company Bloomberg, the analytics company PostHog and Tano. At Tano one engineer owns a feature end to end, and an AI agent reviews every pull request.

• He merges more than 200 pull requests a month at Tano, about eleven times his rate at PostHog, where every change had a human reviewer.
• Human review is optional once an engineer knows the codebase. An AI review agent runs on every pull request, and a stronger agent is available for risky changes.
• When a change breaks production, such as a database migration that blocked writes, the failure is added to the review agent's instructions. The agent then blocks the same change next time.
• The costs he reports include a currency bug a human reviewer would have caught and creators paid twice after a migration. He also reports constant task switching and less knowledge of colleagues' work.

Kakkar presents the trade of human review for speed as a deliberate choice at a small company that can fix errors quickly after release. That a larger team would need an owner for the list of failures its review agent checks is an inference and is not the author's claim.

Derived by Working Surface from the article; more in the Playbook. Source line: In most cases, this automated reviewer is the only review on PRs.

4 November 2026: teams building with AI agents are turning each failure or correction into a permanent check, rule or shared record, rather than fixing one result at a time.