AI Verification Scientist: Lean & Formal Methods

Google DeepMind; Greater London; Hybrid; GBP 129,000 - 186,000; Full time

Job summary

Google DeepMind is seeking a researcher in formal verification and AI-driven proof systems. You will work on developing agents that generate formally verified code and proofs, using Lean and related tools, across languages like C/C++.

We encourage publishing and collaboration across interdisciplinary teams. The role emphasizes safety, ethics, and broad impact, with opportunities to contribute to scalable verification infrastructure and open mathematical benchmarks.

Qualifications

• PhD or equivalent practical experience in programming languages/formal methods.
• 4 years of experience with real-world software verification and bug-finding.
• 3 years of experience with proof assistants (Lean, Coq or similar) or SMT-solvers.

Responsibilities

• Develop and improve AI agents that generate formally verified code, algorithms, and proofs using Lean.
• Formalize the semantics of programming languages (e.g., C/C++) in Lean and build verified static analyses.
• Design experiments evaluating AI-driven proof search against open problems in mathematics and real-world codebases.
• Build infrastructure for applying formal verification at scale, translating code, orchestrating proof search, and integrating with internal tools and models.

Skills

Software verification

Bug-finding

Education

PhD degree in programming languages/formal methods

Tools

Lean

Coq

SMT-Solvers

Job description

Google DeepMind is seeking a researcher in formal verification and AI-driven proof systems. You will work on developing agents that generate formally verified code and proofs, using Lean and related tools, across languages like C/C++.

We encourage publishing and collaboration across interdisciplinary teams. The role emphasizes safety, ethics, and broad impact, with opportunities to contribute to scalable verification infrastructure and open mathematical benchmarks.