Click any tag below to further narrow down your results
Links
Dylan Martin, a PostHog engineer, explains why he left the feature flags team for AI product work and what that shift has taught him, mainly about writing evals and deciding where agents should act versus where humans should see and decide. It's a short personal account of the transition rather than a technical deep dive.
- Much of the skill set carries over: systems thinking, failure modes, and proving correctness still matter, so the move felt smaller than the job postings suggested.
- Evals are the main new skill. Unlike unit tests, they need to accept several good answers to the same request and catch subtle regressions, like counting signup events instead of unique users.
- The hardest open question is interface design: when MCP and skills should replace clicking through a UI, and when a human needs visual previews to make the call, such as seeing a flag's blast radius.
PostHog built its warehouse on DuckDB, then hit limits at scale and is now rebuilding it on a custom Trino engine with its own table format, HogLake. The post explains why DuckDB and DuckLake fell short for customers with very large, complex workloads.
- Single-node DuckDB was hard to run safely in the cloud because PostHog had to size each worker before queries arrived, and predicting CPU and memory per query was difficult.
- DuckLake's portable catalog schema limited optimizations: at PostHog's scale a single-table write could take over three minutes, and over 99% of catalog statistics rows belonged to deleted tables.
- At about 2 PB, roughly 1 trillion rows and 15.3 million snapshots, DuckLake broke in many ways, and PostHog ended up carrying patches in a fork instead of contributing upstream.
Neil Kakkar, a London engineer, compares his output at PostHog (pre-AI) with his current AI-native startup Tano, where he ships about ten times as many pull requests per month. He argues the real gains come from removing old bottlenecks like review queues and cross-team communication, which AI makes less necessary, while creating new problems around context switching and focus.
- PR volume went from roughly 8-20 merged per month at PostHog to 200+ per month at Tano, with much of the gain attributed to agents that build full features in 1-2 hours.
- Mandatory peer review and multi-person feature work were bottlenecks; with AI, one engineer can own a feature end to end from design to backend, which cuts communication overhead.
- The new constraints are context switching from running 3-5 agent threads at once, losing shared team context, and pacing yourself when output is so fast.
PostHog shares concrete lessons from shipping 10 AI products used by 100,000 monthly users, covering when to build AI, how to structure agents, and what fails in practice. The core insight: most AI failures are about reliability and context, not capability, and you need to rethink product architecture for agents rather than bolt them on.
- Most AI product failures come from poor reliability, unclear limits, and insufficient context—not lack of capability. PostHog's install wizard improved activation 2x and conversion 5x because it automated a 2-hour task, not because it was technically impressive.
- Agents need complete tool coverage and flexible abstractions to work. Forcing agents through preset tools or leaving out endpoint coverage (like missing experiment-start when experiment-create exists) makes them abandon tasks and ask users to finish manually.
- Context matters more than reasoning power. PostHog improved wizard effectiveness more by giving agents better context than by adding multi-step planning, subagents, or better models—and model upgrades bulldoze your assumptions anyway, so don't over-invest in harnesses.
- Build shared infrastructure and let product teams own agent features. PostHog's MCP server has 5x more daily users than their AI product because it's easier to maintain and lets teams integrate agents into their own products rather than using a monolithic assistant.
This article discusses the pitfalls of shipping products too quickly, emphasizing that user adoption can't keep pace with rapid releases. It outlines strategies for maintaining product velocity while ensuring users understand and engage with new features.
- Shipping speed has outpaced user adoption, creating a growing backlog of unrecognized features that undermines perceived product quality
- Applying the Theory of Constraints, PostHog focuses on raising user adoption as the bottleneck rather than slowing down development
- Launches should be tiered (major rollouts, strategic upgrades, small tweaks) so marketing effort matches actual significance instead of treating every release as a big announcement
- Feature discovery should be built into the product itself, surfacing relevant tools at the moment tied to specific user actions or milestones rather than relying on external announcements