AI Engineer – Own the Agentic AI Core | Python & TypeScript | Stockholm Startup | €110k + Equity

Fullstack Talents | IT Talent Solutions; Stockholms kommun; On-site; SEK 888,000 - 1,110,000; Full time

Job summary

Fullstack Talents | IT Talent Solutions is seeking an engineer to own the AI core of an agentic systems product. You will design multi-agent architectures, build retrieval pipelines, and refine evaluation datasets to ensure reliable, source-backed results for IP/legal workflows.

You’ll work across Python and TypeScript, influence architecture and engineering standards, and contribute from discovery to deployment in an on-site Stockholm setting.

Qualifications

• Shipping agentic AI systems or LLM-powered features to production.
• Production experience in Python and TypeScript.
• Experience building production retrieval‑augmented generation (RAG) systems.
• Designing LLM evaluations with datasets, metrics and analysis.
• Infrastructure experience with vector databases, distributed compute or cloud orchestration.
• Ability to take an AI feature from concept to deployment.
• Clear ownership of technical decisions and measurable outcomes.
• On-site in Stockholm.

Responsibilities

• Design multi-agent systems for patent research and legal workflows.
• Build retrieval pipelines across patents and scientific literature.
• Develop evaluation datasets and frameworks for accuracy and reliability.
• Improve agent behaviour through context engineering, tool use and state management.
• Own features from customer discovery through deployment and production monitoring.
• Work across Python and TypeScript, including supporting infrastructure.
• Diagnose production issues involving retrieval quality, latency, cost and agent execution.
• Ship frequently and learn directly from customer use.
• Establish practical standards for building, evaluating and releasing AI capabilities.

Skills

Agentic AI

Python

TypeScript

RAG systems

LLM evaluations

Infrastructure

Ownership

Customer value

Tools

Vector databases

Distributed compute

Cloud orchestration

Job description

Employment type: Full-time

Salary guide: Approximately €80,000-€100,000 annually, depending on experience, + equity.

Candidates must be based in Stockholm or willing to relocate for an on-site role.

About the company:

Our client is an early-stage startup building agentic AI for intellectual property and legal teams. Its software helps leading IP firms and Fortune 500 companies analyse large document collections, find supporting evidence and complete complex research more efficiently.

About the role:

Join as one of the first engineers outside the founding team and take ownership of the product’s AI core. You’ll build multi-agent reasoning, retrieval and evaluation systems that help legal professionals work with reliable, source-backed results.

This suits a product-minded engineer who enjoys moving independently from customer problem to deployed capability, with substantial influence over architecture and engineering standards.

What you’ll get to build and influence:

• Design multi-agent systems for patent research, infringement analysis and related legal workflows
• Build retrieval pipelines across large collections of patents and scientific literature
• Develop evaluation datasets and frameworks to measure accuracy, grounding and reliability
• Improve agent behaviour through context engineering, tool use and state management
• Own features from customer discovery through implementation, deployment and production monitoring
• Work across Python and TypeScript, including the infrastructure supporting your systems
• Diagnose production issues involving retrieval quality, latency, cost and agent execution
• Ship frequently and learn directly from customer use
• Establish practical standards for building, evaluating and releasing AI capabilities

What we’re looking for:

• Experience personally shipping agentic AI systems or LLM-powered features to production
• Professional software development experience in both Python and TypeScript
• Practical experience building production retrieval‑augmented generation (RAG) systems
• Experience designing LLM evaluations, including datasets, metrics and analysis of results
• Hands‑on infrastructure experience with vector databases, distributed compute or cloud orchestration
• The ability to take an AI feature from initial specification through deployment
• Strong engineering fundamentals and judgement about customer value
• Clear ownership of technical decisions and measurable outcomes
• Willingness to work on‑site in Stockholm

The strength of your work matters more than a particular number of years. Be ready to explain what you built yourself, how it performed and what you changed when it failed.

Additional experience that would be valuable:

• LegalTech, intellectual property or document‑intensive research products
• Enterprise security requirements, including access controls and data residency
• Supporting SOC 2 compliance
• Early‑stage product development and direct customer collaboration
• Own core AI architecture at an early stage of the business
• Work on substantial reasoning and retrieval problems with real professional users
• Influence both product direction and engineering practices
• Receive salary and equity
• Access unlimited Claude Code for AI‑assisted development and experimentation
• See your work reach customers through frequent production releases

Location:

Stockholm, Sweden. This is an on‑site position.

For questions, please contact: [e-mail]

By replying to this job offer, you agree to our processing of your personal information and confirm that you have had the opportunity to learn about our data handling.