Roadmap
Goals
Missions for 2026
We’ve established the AI Wizard as the fastest way to activate PostHog
This means agentic integration for multiple products across multiple frameworks. This also means support for additional “skills” like migrations and troubleshooting. Ultimately, the Wizard closes the gap between curiosity and real usage. It's an agent that delivers working PostHog solutions within a user’s codebase in minutes.
We’ve created a content-as-context platform that gives us a competitive edge
This means our content like docs, handbook, and blog posts play a central role in context engineering. By structuring our content into composable and reusable context, like skills, we can build agentic workflows that show how the entire product suite works together in practice. Success means developers see our content platform as both a well-organized knowledge base for readers and a context-rich ecosystem for agents.
We’ve built systems that make product docs mostly self-serve
This means each product engineering team writes and maintains their own product docs at a strong baseline level of quality by using the templates, tools, and processes we’ve put in place. We partner with teams to review, improve, or overhaul their docs when needed. Success means keeping current product docs updated and launching several new products with docs, without slowing down.
Goals for Q4 2026
Optimize the wizard as a platform for data-first PostHog setup
Driver:
Edwin Lim
Edwin Lim
Vincent Ge
Vincent Ge
Motivation: The wizard should push adoption of PostHog's data-first products and use cases. We'll optimize the wizard to install and instrument the PostHog platform to deepen data capture and data usage for new users.
What we'll ship:
- Wizard support for all data-first PostHog products in the default wizard command
- MCP install and setup as a primary interface for PostHog usage
- Refreshed TUI, content, and setup flow to showcase value of PostHog's data platform
- Wizard-as-platform improvements in performance, speed, in-app handoffs, and exensibility so other teams can build on top of it
We'll know we're successful when:
- The wizard is improving activation of data-first products and features
- The wizard is increasing MCP installations and usage
- Users reach their first meaningful use of that data faster
- Users praise the wizard's new data-first experience in #brand-mentions
Distribution ideas:
- Position the wizard as a personalized PostHog data setup consultant. Community members are asking for help setting up PostHog in Discord. The wizard should be our first recommendation.
Synergy:
Growth Team
Growth Team is owning the headless CLI wizard that works with anyone's coding agent
Growth Team
Growth Team is owning the in-app cloud wizard for cross-sell, upgrade opportunities, etc.
Forward Deployed Engineering Team
Forward Deployed Engineering Team is owning a first-class migration wizard program for data competitors
Build an agent RAG system for updating and searching our docs
Driver:
John Waters
John Waters
Sarah Sanders
Sarah Sanders
Motivation: The quality of our RAG search for docs is getting worse and so is the performance of our docs writer agent. AI agents now read our docs ~2:1 compared to carbon units (humans), so our docs need better tools and systems to stay correct and up to date for AI consumption.
What we'll ship:
- An integration with our Business Knowledge product to index and search across our docs, repos, and other sources through a shared RAG system
- An agent writer system that creates docs PRs from multiple triggers: monorepo merges, support tickets, docs evals, scouts, Slack tags, feature flag rollouts, etc.
- Improve the agent writer's accuracy, consistency, and adherence to our style guide
- Replace the website AI chat with a new RAG-powered feature
- An approval / notification system for docs PRs generated by the agent writer
We'll know we're successful when:
- Docs PRs merge at the highest velocity ever
- Product teams and developer marketers trust that their product docs are up to date and accurate
- We can reuse the new agent writer to maintain the handbook, READMEs, changelog, and other technical content
- New RAG system provides useful answers from our docs in #at-posthog-anything and on the website
- We're providing great product feedback to the Conversations Teams
Distribution ideas:
- Write a blog post about how we used Business Knowledge to overhaul our docs search and maintenance systems
- Expose our docs RAG search through a public search-docs API endpoint for users and agents
Synergy:
Conversations Team
Conversations Team is building the Business Knowledge product
Support Team
Support Team can update and fix the docs using our system
Measure and improve how well agents use PostHog docs
Driver:
Edwin Lim
Edwin Lim
John Waters
John Waters
Sarah Sanders
Sarah Sanders
Motivation: We want to continually improve how people's agents use the PostHog platform according to our docs. To do that, we'll need a benchmark. We want to measure and evaluate how agents interpret, consume, and execute our product docs.
What we'll ship:
- An automated evals suite that tests models, efforts, tools, and prompts against our docs
- A qualitative or quantitative criteria for specific tasks like install and instrumentation
- Eval configs for each docs page
- A trigger for the docs writer agent to make improvements based on eval results
We'll know we're successful when:
- We can evaluate new model releases and understand how well each use PostHog with and without our docs
- Eval failures lead to docs improvements that increase agent success rates
- We have an official benchmark for the wizard and all PostHog installation / getting started tasks
- We have visibility into how agents use our docs in the wild
Distribution ideas:
- Publish our eval results as a blog post or official benchmark for how well agents can use a product with only its public documentation as context.
Synergy:
Growth Team
Growth Team is running broader evals of agents using PostHog through MCP, CLI, and SDKs.
Turn docs into a skill library for robots
Driver:
Vincent Ge
Vincent Ge
Edwin Lim
Edwin Lim
Motivation: Turn docs into context food for robots. We want to restructure the docs to basically be an instruction manual that's designed for agents. Our docs should be directly usable as micro-skills.
What we'll ship:
- Rewrite and reorganize docs around task-based, reusable micro-skills
- Rethink docs navigation and information architecture
- Build new in-page components for structured prompts, skill downloads, etc.
- Expose docs and skills through a website API for agents and external tools
- Canonical tag and reference system of our docs for agents to use as a map
We'll know we're successful when:
- Most docs pages are packageable as a skill
- There's a feedback loop with both the evals project and RAG / docs writer project
- We can effectively replace the context mill for the wizard
- We see more agent usage and traffic with our docs
Distribution ideas:
- We can use skills as an actionable CTA across blogs, newsletters, and landing pages
- Run a suite of AEO experiments on the docs with
Natalia Amorim
Natalia Amorim - Publish our skills everywhere on open source, like third-party skill aggregators and MCPs like Context 7
Create educational content people love
Driver:
Sarah Sanders
Sarah Sanders Wizard & Docs Team
Wizard & Docs Team
Motivation: We want to make the best educational content for people learning how to build products with PostHog. Our Learn content should focus on teaching humans the technical concepts, practical use cases, and value props of the PostHog platform.
What we'll ship:
- Hands-on learning courses that teach end-to-end product development and data capture with PostHog
- Interactive example apps where readers can explore real data, experiment, and learn by doing
- Reorganize Docs AI and Learn to make educational content easier to discover and navigate
- New data-first content that teaches data capture and usage with PostHog platform
We'll know we're successful when:
- Customer-facing teams regularly share Learn content with users and prospects
- Developers and community members recommend our content because it's genuinely useful, even outside PostHog
- We see organic mentions, positive feedback, and in-person praise for the quality of our technical content
- Other PostHog teams teams re-use our content and components
Distribution ideas:
- Lots to draw from the content playbook: socials, book clubs with community, paid ads, blog, and emails.
Synergy:
- Align with
Developer Marketing Team
Developer Marketing Team on products to prioritize. Create educational content that supports their upcoming marketing launches and campaigns.
Handbook
Slack channel
What we do
- Write and create content that helps humans and robots learn and succeed with PostHog
- Build and maintain the docs knowledge base and platform
- Build and maintain the AI wizard and context mill











