Lean Theorem Proving Engineer, Verified Code Generation, DeepMind

Overview

In this role you will advance formal correctness and security specifications for real-world systems using Lean. You will encode programming language semantics in Lean and design experiments for AI-driven autoformalization and proof search on real-world codebases. You will build scalable infrastructure to translate code, orchestrate proofs, and integrate with Google-internal tools and models. This position sits at the intersection of formal methods and AI, contributing to safety-driven, high-impact AI research.

Pay / Benefits

• 15% bonus target
• equity
• benefits
• career development
• global opportunities

Responsibilities

• Formalize correctness and security specifications in Lean
• Encode language semantics (C/C++/Rust) in Lean
• Design and run experiments on AI-driven autoformalization and proof search
• Build infrastructure for large-scale formal verification, code translation, and proof orchestration
• Collaborate with cross-disciplinary teams and integrate with internal tools and models
• Contribute to safety, ethics and responsible AI initiatives

Key requirements

• PhD or equivalent experience in computer science or mathematics
• 3 years of experience with the Lean proof assistant
• 3 years of experience with using AI agents

• Strong collaboration and teamwork
• Excellent communication of complex ideas
• Research-oriented problem solving
• Lean proof assistant
• AI agents for code or mathematical reasoning tasks
• Formal verification and semantics encoding

…