Own agent runtime and cloud infrastructure: durable execution, Linux computers, model routing, provisioning, updates, recovery, and the security boundaries behind always-on work.
What you'll own
- Build agent kernel primitives for durable sessions, tool execution, context, and model routing.
- Operate per-user Linux runtimes, provisioning, host updates, backups, and recovery.
- Design PostgreSQL persistence, queues, resource limits, and safe concurrency.
- Make tenant isolation, secrets, network security, and observability dependable by design.
What we're looking for
- Experience building and operating distributed systems or developer infrastructure.
- Strong Linux, networking, TypeScript, and PostgreSQL fundamentals.
- Practical experience with agent runtimes or inference integrations and clear failure handling.
- Evidence of thoughtful incident response, security judgment, and tests-first engineering.
What this role offers
- High cash compensation and meaningful equity.
- Deep ownership of architecture and operations, working directly with the founders.
- Frontier AI tools and a product we use ourselves every day.
- Remote work with substantial European-hours overlap.
How to apply
Send a concise application that includes:
- 01Why Matrix and this role?
- 02Describe the hardest system you operated and how you made it reliable.
- 03Share a security or reliability tradeoff you made and what you learned.
- 04Links to code, systems, or technical writing.
Clarity, initiative, and evidence matter more than polish. Please do not send a generic AI-written cover letter.