Context Engineer

Product development used to mean manually writing code, running analysis, diagnosing bugs, and rolling out changes using dozens of tools.

We started with open-source product analytics, launched out of Y Combinator's W20 cohort. We've since shipped more than a dozen products, including:

PostHog Code, the only AI devtool that understands your product, not just your codebase.

Make technologies explainable and legible to both humans and agents
Maintain and design knowledge bases
Engineer and orchestrate context pipelines
Build and ship agents powered by the above
You'll spend your days switching between prose and code – often in the same hour. You’ll create content for humans and context for agents, and the line between those two audiences is disappearing fast.
The work is inherently multidisciplinary. From our experience, the people who are good at context engineering tend to be natural teachers or strong writers who can code at a high level. There's more liberal arts to building robots than you'd might expect.

This role combines technical content and AI engineering. You'd be a Context Engineer on the Wizard & Docs team.
Excel at teaching and communication. You’re great at educating people by explaining technology in a hands-on and illustrative way. Your preferred medium might be writing, building apps, making visuals, or recording video. Whatever it is, you can take a complex system, teach it clearly, and then build it.
Have real development chops. You have full-stack software development experience either in full-time roles, side projects, open source, whatever. We welcome all backgrounds, but the technical bar is high. We run production services, ship PRs across multiple codebases and infrastructure, manage a support queue, and get tagged by engineers for code review.
Are multidisciplinary. You think laterally and enjoy approaching problems from different angles, switching hats as the work demands: writing docs, shipping code, organizing IA, creating diagrams, building agents or CI/CD pipelines. You're looking for a role that lets you flex all these muscles, where a wide skillset is valued.
Think in systems. You won't just write a doc and move on. You'll think about how that doc gets turned into a skill, how an agent executes it, and what happens when the underlying product changes. The whole pipeline is your problem.
Build agents and AI tools. You already have experience building on top of LLMs and have strong opinions about what works with regards to the context window, harnesses, coding agents, and more.
Bias to action. We ship fast (like, weirdly fast). There's rarely a blueprint to follow, or a list of jobs to be done here. You need to be comfortable being reactive, iterating quickly, and shipping PRs as priorities change.
We’re building a self-driving AI product on top of a large collection of dev tools, so the surface area of problems we help solve is huge.
This means you’ll be:
Working on our AI Wizard, an AI agent that onboards thousands of users to PostHog
Building custom agents, harnesses, skills, context pipelines, and MCPs and shipping them to production
Building systems to maintain our developer docs and knowledge bases using docs-as-code and AI engineering techniques
Creating technical content across multiple mediums like docs, example apps, or video explainers
Creating workflows for publishing streams like the changelog
Working with other teams so they can build on top of our systems
AI Wizard – an AI onboarding agent that onboards all users to PostHog
Writer agent – a RAG powered agent system that maintains our docs
Context mill – a skill assembly and context delivery service
Developer docs – the company’s knowledge base
❌ Gathering specs and interviewing SMEs to write reference docs
❌ Running webinars. Eugh, just no
❌ Writing newsletters and brand content
❌ Waiting to be told what to do. No one will hand you a list of tasks or projects to build
A technical writer who regularly ships code and builds features
A developer advocate who builds as much as they write
A software engineer who’s looking to change track
A solutions or growth engineer who enjoys creating educational content
An instructional or content designer who is technically gifted
Or something else, but you have these traits and skills
We're a team of ex-technical writers, ex-software engineers, ex-devrel, ex-PMs, and a bunch of other past lives. We ditched the old titles because the work doesn't map to any of them. We call ourselves context engineers now, and we're still figuring out what that means.
Software development experience. It doesn’t have to be formal, but you'll be writing production code, not just code samples
Proficiency working with LLMs, coding agents, MCPs, and the latest AI engineering tools
A portfolio of technical content (personal website, GitHub profile, YouTube channel, zip file, etc.)
Excellent writing and communication skills
Experience creating educational content for developers
An opinionated take on what makes great docs for humans and AI
#LI-DNI