Founding AI engineer
Founding AI Engineer — Opengeni (on site, Norway, you need to have working visa to apply)
We’re looking for a hands-on Founding AI Engineer to help turn this vision into reliable, useful software.
You’ll work directly with the founders, shaping both the product and its technical foundations. This is a product engineering role focused on making agents work in practice—from understanding a task and using tools to recovering from failures and delivering a verifiable result.
What you’ll build
- Agent execution: Systems for planning, tool use, long-running tasks, error recovery and human approvals.
- Software development workflows: Agents that work with repositories, issues, tests and delivery pipelines.
- Context and knowledge: Ways for agents to retain useful context and apply shared knowledge across tasks.
- Evaluation and reliability: Tests, tracing and feedback loops that measure whether agents actually complete work correctly.
- Integrations and infrastructure: Connections to real tools, secure execution environments and maintainable self-hosted deployments.
- The product itself: Features taken from an ambiguous user need through implementation, release and iteration.
What we’re looking for
- Strong software engineering fundamentals and experience shipping software people use.
- Hands-on experience building with LLMs, agents or tool-calling systems beyond a basic chatbot.
- The ability to debug across model behavior, application code and infrastructure.
- Good judgment about when to use an agent and when conventional software is the better solution.
- Comfort taking ownership of loosely defined problems and making practical trade-offs.
- Clear communication and an interest in working closely with users.
- A genuine interest in open source and giving people ownership of their tools.
Experience with developer tooling, distributed systems, sandboxed execution, agent evaluations or open-source maintenance would be especially relevant.
Why join
- Help define a product and its architecture at an early stage.
- Work on the full problem of AI execution, not only prompting or model integration.
- Build openly, with work that other developers can inspect, use and extend.
- Start with a concrete problem in software development while contributing to a broader vision for how work gets done.
Apply
Send us your GitHub, a project you’re proud of, or a short description of something you’ve built. Tell us what you owned, what was difficult and what you learned.