Research Engineer, Foundation Model

Prior Labs GmbH; North Merrick (NY); Hybrid; USD 150,000 - 210,000; Full time

Job summary

Prior Labs GmbH seeks Research Engineers to advance TabPFN technology. You will design experiments, implement training pipelines, and contribute to papers while building scalable ML systems in an elite team. Offices in Europe and New York support collaboration, with remote work possible in exceptional cases.

You should have a Master’s or PhD in CS, strong Python/PyTorch skills, and published research. Experience with transformers or foundation models is preferred and a passion for efficient,

Qualifications

• Master's or PhD in Computer Science or related field.
• Publications at top ML venues or equivalent demonstrated research impact.
• Proficiency in Python, PyTorch, and ML ecosystem (scikit-learn, pandas, NumPy).
• Experience implementing and training neural network architectures, ideally transformers or foundation models.
• Understanding training dynamics, scaling behavior, and failure modes in DL systems.
• Interest in model efficiency and deployment-ready optimization.

Responsibilities

• Design experiments and ablations for TabPFN scaling and architectures.
• Contribute to papers and code that turn architectural ideas into trained models.
• Own training and evaluation infrastructure to enable rapid experimentation.
• Maintain high software quality alongside research deliverables.
• Collaborate with researchers and engineers across the team.

Skills

Research experience

Publications

Python

PyTorch

scikit-learn

Pandas

NumPy

ML systems

Training dynamics

Model efficiency

Education

Master's or PhD in Computer Science

Tools

Python

PyTorch

Scikit-learn

Pandas

NumPy

Job description

Who we are

Foundation models transformed text and images. Structured data - the largest and most consequential data format in the world - stayed untouched, until now. What LLMs did for language, we're doing for tables.

We pioneered tabular foundation models: TabPFN v2 was a Nature cover story, has passed 3.5M+ downloads and 7,500+ GitHub stars, and runs in production from detecting lung disease with Oxford Cancer Analytics to preventing train failures with Hitachi. The hardest problems - millions of rows, real-time inference, entirely new modalities - are still open, and no one else is working on them at this level.

We're a small, highly selective team of 40+ with backgrounds from Google, DeepMind, Meta, Apple, Amazon, Jane Street, and CERN, led by Frank Hutter, Noah Hollmann, and Sauraj Gambhir, and advised by Bernhard Schölkopf and Turing Award winner Yann LeCun.

In July 2026, less than 18 months after our €9M pre-seed, we joined SAP as an independent frontier AI lab - same team, mission, and open-weights models, now backed by more than €1 billion over four years.

About the role

Tabular data breaks the assumptions that make scaling work for language and vision. There's no natural sequence, no spatial structure, no shared vocabulary across datasets. The architectures and scaling laws that power LLMs don't transfer. We've made the first breakthrough with TabPFN - the hardest problems are still ahead.

At Prior Labs, Research Engineers are the science team. You'll design experiments, contribute to papers, and write the code that turns architectural ideas into trained models - the same people do the research and the engineering, which is why both are good. You'll have significant technical ownership and room to grow as we scale.

The problems we're solving:

• Scaling transformer architectures from 10K to 1M+ samples - without the structural assumptions that make language models scale

• Building multimodal models that combine tabular, text, and numerical understanding

• Making models efficient enough for real-world deployment, not just accurate enough for a paper

• Designing architectures for time series, forecasting, anomaly detection, and multiple related tables

Day-to-day, you'll design and test novel architectures, run ablations, analyze scaling behavior, and write the training and evaluation infrastructure that makes rapid experimentation possible. We hold software quality to the same standard as research quality.

What we're looking for

• Master's or PhD in Computer Science or a related field, plus 3+ years of experience building ML systems in research or industry

• Publications at top ML venues (NeurIPS, ICML, ICLR, etc.) or equivalent demonstrated research impact (widely used open-source, deployed systems)

• Deep proficiency in Python, PyTorch, and the broader ML and data science ecosystem (scikit-learn, pandas, NumPy), with strong software engineering practices

• Experience implementing and training neural network architectures, ideally transformers or foundation models

• Solid understanding of training dynamics, scaling behavior, and common failure modes in deep learning systems

• Genuine interest in model efficiency - making large models faster, more scalable, and practical to deploy

Nice to have

• Experience at an early-stage startup or as a founding engineer

• Contributions to open-source ML libraries or tools

• Experience with model distillation, inference optimization, or on-device ML

• Background in tabular data, time series, or other structured data - helpful but not required

Life at Prior Labs

You'll work alongside researchers and builders who hold themselves to a very high bar - in the quality of their work and in how they work with each other. We move fast and still take the time to do things right.

Our teams are based in Berlin, Freiburg, and New York - when you're working on something as hard as TabPFN, being in the same room matters. But great people come from everywhere, and in exceptional cases we're open to remote, which usually means frequent travel to one of our offices. Wherever you're based, the whole company comes together regularly for offsites to build and celebrate together.

Our Commitments

The best products and teams are built by people with a wide range of perspectives and backgrounds. We welcome applications from all identities and walks of life - especially if you've ever felt discouraged by "not checking every box" - and provide equal opportunities regardless of gender, sexual orientation, origin, disability, or any other trait that makes you who you are.

We care about how your data is handled - see our Recruiting Data Privacy page