GTM Engineer - Kitaru & ZenML

Overview: The Person Who Turns Open Source into Design Partners

We're hiring a GTM Engineer to own the motion that turns curious developers into design partners - and design partners into the people who tell us what Kitaru and ZenML should become next.

One-line version: someone who can whiteboard an integration architecture in the morning, pitch it to a Head of AI in the afternoon, write the blog post that explains it on the train home, and show up at a meetup that weekend to talk about it. Technical enough that engineers trust them. Commercial enough that founders take their meetings.

You'll work across our two open-source products: ZenML, the AI platform thousands of teams use to orchestrate ML pipelines and agent workflows, and Kitaru, our brand-new durable execution layer for Python agents - crash recovery, human-in-the-loop, replay from any checkpoint. ZenML has years of production trust and a warm community giving us a steady stream of signals. Kitaru just launched, has no playbook yet, and is defining a category. That's where you'll spend most of your week.

This is not a sales role. It's not a marketing role. It's not exactly a DevRel role either. It's the role that creates new categories by finding the first ten companies willing to co-build the product with you, shipping the integrations and content that get them in the door, and feeding everything you learn straight back into product strategy. If "technical GTM" is a thing, this is the job.

We expect you to be AI-pilled. You're not just building for AI engineers, you are one. You have strong opinions about which agent frameworks ship and which won't survive 2027. You can defend those opinions in a room of engineers who'll push back. But AI-pilled ≠ vibes-pilled - you read every diff your agent produces, you can architect a real system on a whiteboard, and your judgment is yours.

Key Responsibilities (The "Jobs to be Done")

• Hunt and Land Design Partners: Identify the top companies who most need Kitaru's durable execution or ZenML's pipelines, get a real conversation with the right person, and turn it into a working POC. This is outbound, targeted, founder-style hustle - not inbound qualification. You should be excited about cold-emailing a CTO with a thesis they haven't heard yet.
• Co-Build with Customers: Once design partners are on board, you're their internal champion. You architect the integration with them on the whiteboard, ship the first demo yourself, write the case study afterward, and make sure their Slack questions are answered same-day.
• Ship the Integrations That Get Us In the Door: Build integration code for the frameworks our users live in - PydanticAI, OpenAI Agents SDK, Claude Agents SDK, LangGraph, LlamaIndex, CrewAI, memory layers like cognee, vector stores, graph stores. Each integration is both a product and a marketing asset. You write the code, the example app, and the launch post.
• Shape Where Kitaru Goes Next: Sit in product strategy conversations as a peer, not a stakeholder. The product vision for Kitaru is being written right now - your customer conversations and integration directly inform what we build. You'll write spec docs, argue with the founders, and own pieces of the roadmap.
• Own the Technical Narrative: Write the blog posts, give the meetup talks, host the workshops. Not generic content - the specific thing that explains why durable execution matters to someone debugging a 3,000-step agent run at midnight. Both real technical depth and a sense of when to stop being technical.
• Live in the Community: Slack, GitHub, Discord, X, Hacker News, AI meetups across Europe and beyond. You spot the user asking the question that's actually a Kitaru-shaped problem and reply within the hour. You're the human face of the product in places where the product doesn't speak for itself yet.
• Build the GTM Playbook as You Go: There isn't one yet for Kitaru. You'll figure out what outreach lands, which channels convert, what design-partner profiles work, what content compounds, and document it so the next hire after you isn't starting from zero.

Tech You'll Work With

• Our Products: ZenML (pipelines, stacks, snapshots, deployer, control plane) and Kitaru (durable flows, checkpoints, replay, human-in-the-loop).
• The Code You'll Write: Python 3.11+ (decorators, async, type hints - IC-level competence; you don't need to be a production-shipping engineer, but you do need to write code that runs). Enough TypeScript to update a docs site or example app when it matters.
• The AI/Agent Ecosystem: PydanticAI, OpenAI Agents SDK, Claude Agents SDK, LangGraph, LlamaIndex, CrewAI. Vector stores, graph stores, memory layers. You know the landscape well enough to credibly position us inside it.
• The GTM Stack: Attio, GitHub Projects, Slack Connect with customers, GitHub for both code and conversation, LinkedIn for outreach.