Reading up on PostHog

Engineers can offload AI‑generated code review to agent pipelines, using multiple specialized reviewers and automated triage to reduce human bottlenecks.

An experiment using Claude Code to autonomously optimize a file compression algorithm over 10 iterations, achieving competitive results against standard tools by iterating on an LZSS-based approach.

PostHog's engineer used parallel Claude Code sessions to rewrite their SQL parser, achieving a 70x speedup by fuzzing against the old ANTLR-based parser as an oracle.

The author used AI agents to rewrite PostHog's SQL parser in Rust, achieving a 70x speedup (454x in production) via property-based testing and oracle-based development.

Loops—self-prompting agents that autonomously complete long-running tasks using context and verification—are becoming viable due to model improvements and enable self-driving products.

PostHog used a Karpathy-style AI agent to find a 3-year-old ClickHouse bug where `toTimeZone()` disabled primary key usage, and the fix cut query time by 37%

PostHog Code uses production data signals to automatically diagnose issues and generate pull requests before users know there's a problem.

PostHog shares lessons from two years building AI agents, urging teams to validate demand with simpler interfaces like MCP servers before committing to custom agents.

PostHog argues that rapid product velocity creates an adoption bottleneck where user attention, not shipping capacity, becomes the limiting factor, and offers strategies to balance both.

AI coding tools like Claude Code reduce the need for deep flow states, freeing mental capacity to enjoy spontaneous meetings and interruptions.