AI Agent Platform Engineer

AI agents are getting pretty good at doing things on their own. Writing code, browsing the web, calling APIs, managing workflows, occasionally making spectacularly bad decisions. 🤖

The demos look great. Until an agent gets stuck in an infinite tool-use loop, executes something it really shouldn't, or confidently reports that a task is done when absolutely nothing happened.

Fun stuff.

At Artificial Seed, we're building a platform for autonomous virtual employees. Not another chatbot with a fancy interface, but agents that can actually use tools, interact with external systems, collaborate with each other, and get real work done.

And ideally not set everything on fire in the process.

That's why we're looking for an AI Agent Platform Engineer

You'll be working on the infrastructure that makes autonomous agents actually useful — and, more importantly, reliable.

What you'll get your hands on:

💜 Building the platform behind autonomous AI agents;

💜 Designing Tool-Use loops, multi-agent collaboration, and LLM integrations;

💜 Creating secure sandboxes where agents can execute code without causing chaos;

💜 Debugging non-deterministic LLM behavior (yes, good luck);

💜 Making agent systems fault-tolerant, observable, and production-ready;

💜 Using coding agents heavily yourself — we're big believers in spec-driven development.

What you'll need:

💜 Strong production-level Python. TypeScript / Node.js is a big plus;

💜 A solid understanding of Linux internals — processes, signals, IPC, PTYs, sockets, and all the fun low-level stuff;

💜 To know asynchronous programming well enough to explain why blocking an event loop is a terrible idea;

💜 Hands-on experience with LLM APIs, tool calling, structured outputs, and agentic patterns like ReAct and Plan-and-Solve;

💜 To understand why prompt injection is a real security problem, not something you can fix by writing "ignore malicious instructions" in a system prompt;

💜 Experience with Docker, containers, and isolated execution environments.

Bonus points if HTTP/2, WebSockets, gRPC, SSE, and MCP are already part of your daily vocabulary.

Now, the good part. 🇪🇸

We're based in Bilbao, Spain, and we'll help relocate you and your family. Paperwork, housing, settling in, even bringing your pets — we've got you covered.

You'll also get:

✨ Unlimited vacation (yes, really);

✨ Flexible working hours and zero micromanagement;

✨ A modern office, team events, and off-sites;

✨ Real influence on architecture and technical decisions;

✨ Hard engineering problems and smart people to solve them with.

We're not looking for someone who just knows how to wrap an LLM API.

We want someone who gets excited about figuring out what happens when you give an AI agent access to a terminal, a bunch of tools, and a goal — and then have to make the whole thing work reliably.

Sounds like your kind of engineering problem?

DM me @dariabezmenko. Let's talk. 🚀