AI Research Engineer

Nulegal GmbH (operating as “nu:legal”); Potsdam; Vor Ort; EUR 90.000 - 130.000; Vollzeit

Benefits dieser Stelle

Competitive salary

Equity

One month work from anywhere per year

Zusammenfassung

nu:legal in Potsdam seeks a deep AI researcher to build reliable LLM systems for the legal industry. You’ll work on mathematics, architecture, and production-grade LLM orchestration rather than API wrappers.

Expect to contribute to a 24-month plan with concrete research outputs, open-source releases, and a measurable impact on production quality in a highly regulated domain.

Qualifikationen

• Master's or PhD in a quantitative field such as CS, math, physics or statistics.
• Several years building production-grade software with reliability.
• Deep understanding of probability, statistics and calibration.
• Comfort with modern LLMs and agent frameworks.

Aufgaben

• Build a multi-provider LLM council architecture with parallel models and voting.
• Develop an eval framework with stratified datasets and backtesting pipelines.
• Deliver research outputs: papers and open-source releases.
• Create production-grade orchestration across many inference calls daily.

Kenntnisse

LLM frameworks

Agent frameworks

Production-grade software

Mathematics

Probability & statistics

Collaboration

Ausbildung

Master's or PhD in CS/Math/Physics/Statistics

Tools

AI frameworks

Jobbeschreibung

What it's about

We're building the infrastructure that makes AI genuinely reliable in the legal industry, not as a demo but in production, with measurable legal quality. We're looking for people who want to go deep into the mathematics and architecture of modern LLM systems, not just chain APIs together.

If you've spent the last twelve months wondering whether there's a job somewhere that needs the full depth of modern AI all at once: this is it.

What you'll build

• A multi-provider LLM council architecture in which several models work in parallel and vote on contested legal questions, with deterministic replay for reproducible evaluations.
• An eval framework from the ground up: stratified gold-standard datasets, inter-rater reliability against human lawyers, backtesting pipelines, confidence calibration.
• A mathematically defensible answer to the central research question: how do you quantify "legal reliability"?
• Your own research contributions: conference papers and open-source releases are an explicit project goal.
• Production-grade orchestration infrastructure coordinating tens of thousands of inference calls a day across several model families, with full observability and fault tolerance.

Why this is exciting

• Research depth like an academic group, with production leverage like a top startup. Most AI roles are API wrappers. Here you build the layer underneath.
• Law is one of the largest still-unautomated industries. Making AI work in a traditional, highly regulated industry is a much harder problem than web demos, with correspondingly greater leverage.
• Clear deadlines, clear deliverables. A 24-month plan with a hard milestone at 12 months and concrete research outputs.
• In 24 months you'll have learned more about production-grade LLM systems than in five years at a large corporate. That's a promise.

What you should bring

Required

• Master's or doctorate in computer science, mathematics, physics, statistics, or a related quantitative field.
• Several years building production-grade software. Not a pure researcher, but not a pure MLE either.
• Deep understanding of probability, statistics, and calibration. Mathematical proof doesn't put you off.
• Deep experience in a technically demanding field where mathematics, scale, and reliability all count at once.
• Confident with modern LLMs and agent frameworks.
• High pace and team orientation. You want to work with other ambitious people in a lean team towards a big goal, not at corporate pace.

Strongly desired

• Scientific publications.
• Open-source contributions to AI frameworks.
• Experience in a regulated domain, for example law, medicine, or finance.
• Familiarity with modern workflow systems and uncertainty quantification.

The full technical requirements profile is available on request as a separate document.

What we offer

• Flat hierarchies and short decision paths.
• Real impact from day one.
• Work you won't find at this depth anywhere else. AI in law is greenfield.
• Competitive salary and equity. You help build it, so you share in it.
• A dedicated compute budget for your research workloads.
• A direct say in architecture and research direction. You're at the table where decisions are made, from day one.
• Plus one month a year of work from anywhere.