AI Engineer (Agentic AI / Full-Stack): Founding Engineer at a Stealth AI lab

Stealth Startup; Stockholms kommun; Hybrid; SEK 332,000 - 443,000; Part time

Job summary

Stealth Startup in Stockholm is seeking an AI Engineer to build production-grade, agentic AI workflows for regulated fintech/insurtech clients. You'll work from backend to frontend, shaping the product roadmap and collaborating with client teams.

We value depth over boxes; you’ll document work, drive cycles daily, and help teams scale reliable systems while navigating governance, security, and compliance in a regulated environment.

Qualifications

• Experience deploying agentic AI in production for regulated environments.
• Proficiency in Python/TypeScript and cloud compute.
• Ability to document and explain technical work clearly.
• Familiarity with privacy, security, governance in fintech/insurtech.
• Willingness to work from Stockholm and travel to client sites.

Responsibilities

• Build AI workflows for client sales teams, from backend to frontend.
• Contribute to core processes, validate with clients before productization.
• Document work and explain it to stakeholders.
• Collaborate with client developers and mentor on the way.
• Operate in a consulting contract setup, with part-time hours.

Skills

Agentic AI

Full-stack

Reliability

Enterprise architecture

Regulated environment

Cloud certs

Tools

AWS Bedrock

Claude certification

Job description

We're a stealth AI lab in fintech, wealth and insurtech, with a strong presence in the Nordics. We're looking for an AI Engineer to build agentic systems for regulated businesses. We don't build demos or mock-ups. We build production systems that hold up under scrutiny in a regulated industry going through major change.

The role

You'll work closely with one of our clients, building AI workflows their sales team actually uses, from backend to frontend. We start with the core processes and close the loop with the client and evals before we build the cool stuff. Whatever keeps coming back becomes product. We already have a tech roadmap, PRD/MRD and product vision, and you'll help shape them.

What you bring

• Agentic AI in production: Claude/Codex, tool calling, agents and agent harnesses, MCP, Skills, RAG, orchestration including multi-agent flows
• Full-stack: Python/TypeScript, backend, databases, cloud, and frontend when needed
• Reliability: evals, benchmarks, guardrails, grounding, structured outputs, observability
• Enterprise architecture: scalability, latency, cost, model abstraction, human-in-the-loop, integration with existing systems
• Regulated environments: AI that is traceable, explainable, secure and governable. Data protection, access control, governance
• AWS/Bedrock and Claude certification or fluency are a big plus

You don't need to tick every box, because we value depth over checklists. What matters is that you're curious about how agentic AI is changing the way software gets built, and you want to be at the front of it.

How you work

• Shared repo from day one. You show what you've built as you go and can explain what it does
• You document as a matter of course
• You close a cycle every day. Momentum over perfection
• You drive things forward without micromanagement, and you coach a couple of the client's developers along the way

The setup, plainly

It starts as a consulting contract / projekt employent, at least half-time, with market-rate pay tied to delivery. You'll be based in Stockholm, with periods on-site at the client. If it works, we'll talk about a founding engineer or technical partner role.

P.S. We're also hiring a SecDevOps / AI Ops Engineer and a Forward Deployed Engineer to guide the client's team through the pilot.