We’re looking for an AI Developer who doesn’t enjoy doing the same work twice. Not someone who avoids effort, someone who avoids unnecessary effort. If for you, AI means more than calling an API, if it means designing systems, orchestrating workflows, and making things actually work, keep reading!
The role
You’ll work on systems where AI is one component. Not the entire job, just one part of it.
This role includes:
- building AI-powered features and agentic workflows
- integrating AI into real products and business processes
- working on backend logic, data flows, and system design
- contributing to AI-Driven Development practices across projects
Because AI is useful. But only if the rest of the system makes sense.
AI-Driven Development (Required)
AI tools are not a bonus. They are part of the workflow. Tools like Cursor, Claude Code (or similar) are expected to be used daily to:
- speed up prototyping
- reduce repetitive work
- focus on system design instead of boilerplate
If you prefer writing everything manually, this role will feel unnecessarily difficult.
Tech stack & what you’ll do
The core stack is:
- Python 3.12+
- Asyncio (used seriously, not just imported)
- strict typing
- Pydantic & Pydantic AI
- MCP (Model Context Protocol) to connect LLMs with tools and data
You will:
- develop AI agents and RAG-based systems
- build MCP connectors and servers to integrate models with tools and data
- design prompts and ensure outputs are structured and validated
- work on data flows (including streaming and complex JSON) toward frontend applications
What we’re looking for
You don’t need to know everything. But you should:
- be comfortable with asynchronous Python
- have hands-on experience (even personal projects) with Pydantic / Pydantic AI
- be genuinely curious about AI systems
- understand the difference between ReAct prompting, Tool Calling, and Function Calling
- English proficiency
No legacy code. No unnecessary process. No pretending things are more complex than they are. Focus: build systems that don’t need constant fixing.
Junior profiles are welcome, especially if motivated to learn quickly. Mid-level candidates should already have a good sense of what works and what usually doesn’t.
Soft skills
- flexibility and adaptability
- willingness to learn continuously
- ability to work both independently and in a team
- comfort with evolving requirements and fast iterations
Nice to have
- Vercel AI SDK
- React / TypeScript
- experience with SSE streaming
- handling typed outputs in frontend environments
- a full-stack mindset (or willingness to develop one)
Compensation & Benefits
Full-Remote.
MacBook Pro and meal vouchers.
Contract type and compensation will depend on experience.
Apply
Send your GitHub or talk about a recent AI project (preferably involving Pydantic AI). No slides. No long introductions. Just code and technical thinking.