Postdoc: Machine-Verified LLM Inference (Keystone Project)
EPFL · Lausanne
Job description
About the role
The Keystone project, funded by ARIA and a collaboration between EPFL and Imperial College London, aims to build a formally verified machine learning inference engine. It seeks a postdoc working in formal methods and/or computer systems, with interests in machine-checked verification of systems software, GPU kernel semantics and verification, and AI-assisted proof engineering. The precise research focus will be discussed with the successful candidate.
Key responsibilities
- Conduct research on the Keystone project, including contributing to the design and construction of a machine-verified LLM inference engine
- Formalize GPU kernel semantics (e.g. PTX/Triton) in a proof assistant and verify high-performance inference kernels
- Implement and verify the inference coordination layer (batching, KV-cache management, scheduling) in Rocq/Lean, with extraction to executable code
- Develop agentic AI workflows for specification autoformalization, proof generation and proof repair
- Build a network in formal verification, systems and ML infrastructure, collaborating regularly with the Imperial College London team (visits to London expected)
- Contribute to open-source releases of specifications, proofs and verified artefacts
- Take part in ARIA activities such as red/blue team exercises and sprint reviews
Required profile
- PhD (or nearly complete) in computer science or a closely related field
- Background in formal verification, programming languages, systems or ML
- Research experience in interactive theorem proving (Rocq, Lean, HOL, Isabelle or similar), GPU programming or semantics, compilers, concurrency or ML systems
- Publication record relative to career stage in leading journals and conferences
- Excellent written and oral English, independent and creative, good team spirit
Required skills
- Strong computational and analytical skills, including software engineering across several languages (e.g. Python, C++, OCaml, functional languages)
- Experience with or evaluation of LLM-based tools for code or proof generation is a plus
What we offer
- A stimulating, international working environment and excellent working conditions
- State-of-the-art research in a high-profile ARIA-funded project, with access to frontier AI models and high-performance compute
- Interaction with experts at EPFL and Imperial College London and a team of postdocs and PhD students
- Funded travel between Lausanne and London and to conferences
Applications are submitted only through the online platform, with a cover letter, CV with publications, research statement and three referee contacts. The contract starts on 1 December 2026 or later, is fixed-term for one year and renewable, with full-time activity.
Questions fréquentes
Why are you reporting this job?
Explore further
Salaries, guides and searches in Switzerland.
Apply in 30 seconds
Enter your email to apply. An account will be created automatically.
By continuing, you accept our terms of use.
Already have an account? Login
A question about this job?
Ask it here: you will get the full job summary by e-mail, right away.
Published 2 weeks ago
Expires 1 month from now
27 views · 0 interested
Boost your chances
Upload your CV — we will match you with relevant openings.
Analyzing your CV...
EPFL
Lausanne
Related job offers
-
Cybersecurity Intern – AI Threats & Cloud Hardening
swissborg Lausanne -
Machine Learning Engineer
EPFL Lausanne -
Cybersecurity Intern
SwissBorg Lausanne -
Application Manager·e (H/F) – Gestion des intégrations et des API
CH Media Aarau -
Tech Lead – Responsable de l’Architecture Technique (100 % CDI)
Touring Club Suisse (TCS) GE