PostHog · July 2026
25 tracked updates · July 2026
In July 2026, PostHog shipped 25 tracked updates. The dominant theme was feature updates (5). Most of the activity landed in the week of Jul 20, 2026. The standout: “Give your agents thefull context” (Jul 31, 2026).
Browse other months from PostHog
PostHog introduced a 'context warehouse' designed to provide AI agents with clean, consistent data without requiring pipelines. The feature combines ELT, managed warehouse, CDP, and reverse ETL capabi…

PostHog introduced Replay Vision, an AI layer for session recordings that automatically detects and fixes user struggles by drafting PRs for review. It identified and resolved two UX issues in PostHog…
PostHog’s marketing roles are highly competitive, with over 715 applications per opening and hiring managers spending 350+ hours interviewing annually. The company prioritizes candidates who ship fixe…
Pricing updated for Posthog: - Scale: Custom pricing → $0/mo - New tier: Pay-as-you-go ($0/mo) - New tier: Enterprise (Custom pricing)
PostHog increased monthly PRs from 1,441 to 4,725 in six months with only 10% headcount growth, aiming for 10,000 PRs/month by September 2026. The acceleration stems from agentic automation, self-driv…
PostHog published a framework for delegating tasks to AI agents based on two factors: ease of checking work and cost of undoing mistakes. They introduced four autonomy levels (0-3) and launched Scouts…
PostHog introduced 'self-driving' agents that autonomously analyze product data, generate pull requests for fixes, and merge changes after review. The system uses signal sources like error tracking an…
PostHog is shifting its positioning from open-source product analytics to a self-driving software product platform, emphasizing building with PostHog rather than adding it later. The company is overha…
PostHog introduced the concept of a 'context warehouse'—a unified tool combining data storage, ingestion, modeling, and direct agent access—designed for AI agents and product teams. Unlike traditional…
PostHog argues that by 2030, AI agents will be primary software users, shifting human roles to judgment, trust, and ambiguity resolution. They propose building agent-first infrastructure with complete…
PostHog introduced distributed tracing in beta, complementing its existing debugging tools (Error Tracking, Session Replay, Logs). Traces provide end-to-end request journeys with spans, waterfall view…
PostHog launched a self-driving feature that uses AI agents to monitor product data, identify patterns, draft fixes, and open pull requests for review. The system operates in a continuous loop, with a…
PostHog introduced Replay Vision, an AI layer over session replays that scans user sessions, classifies struggles, and autonomously opens draft pull requests for bugs detected. In two weeks, scanners …
PostHog details four AI-driven workflows to reduce code review bottlenecks, including agent-based reviews, PR babysitting loops, auto-stamping for low-risk PRs, and observation-based verification. The…
PostHog introduced a provisioning API enabling automated account creation and embedded analytics for end users without requiring manual sign-ups. The API uses OAuth with PKCE, provisions projects, and…
Track PostHog on autopilot