Engineering

AI Agent Engineer

Shape the agent that turns a prompt into a production-ready app: tools, context, and guardrails.

Remote
Full-time
Engineering

Short version - You will shape the agent that turns a builder prompt into a working app: tools, context, planning, memory, guardrails, and evaluation loops.

Why Mythos?

The Mythos product only works if the agent can reason clearly while still feeling fast and direct to the person building.

You will work where model behavior, product UX, and production reliability meet.

What you will shape

  • Design agent tools, prompt contracts, context retrieval, and state transitions for app generation workflows.

  • Build evals and traces that make model behavior debuggable instead of mysterious.

  • Improve tool calling, file editing, preview repair, and deployment handoffs.

  • Turn failure modes into product surfaces users can understand and recover from.

What makes you strong here

  • You have built production LLM systems with real constraints, not just demos.

  • You think in loops: observe, plan, act, verify, and repair.

  • You are pragmatic about models and ruthless about user-visible quality.

  • You can work across backend, frontend, infrastructure, and product language.

How we hire

  • A short intro call focused on fit, motivation, and the work you want to do.

  • A practical craft conversation with the people closest to the role.

  • A work sample or paid trial when it is useful to see the collaboration style in motion.

  • A final founder conversation about ownership, pace, and how you want to shape Mythos.

About your application

Send whatever shows your best relevant work: shipped product, code, writing, prototypes, docs, or a short note about what you want to build here.

We read every application directly. No puzzle hoops, no automated rejection theatre, and no need to overproduce a cover letter.

Ready to build this with us?

Help us make software creation feel calmer, faster, and more exact.

Apply for this role