Research Engineer – Formal Methods / Verification

Acceler8 Talent; San Francisco (CA); On-site; USD 120,000 - 180,000; Full time

Job summary

Acceler8 Talent seeks a Research Engineer to bring formal methods and rigorous reasoning to frontier AI alignment research. You will apply ideas from formal verification, program analysis, and compilers to understand and constrain model behavior, building practical research systems from scratch.

PhD strongly preferred, but deep technical creativity matters most. The environment is small, experimental, and highly technical, offering room to shape a new direction in AI safety and interpretability.

Qualifications

• Strong background in compilers or systems engineering.
• Experience with formal verification or theorem proving.
• Security research or operating systems exposure.
• Research engineering with strong abstraction skills.

Responsibilities

• Apply formal verification and program analysis to model internals.
• Build structured reasoning frameworks for interpretability research.
• Explore compiler-style abstractions for neural computations.
• Develop tooling that combines interpretability with verification-inspired constraints.
• Shape an entirely new research direction inside the lab.

Skills

Formal methods

Compiler/Systems engineering

Verification / theorem proving

Abstraction in research engineering

Education

PhD strongly preferred

Job description

Research Engineer – Formal Methods & Reasoning

An AI research lab focused on alignment, interpretability, and reinforcement learning is hiring engineers to explore how ideas from formal methods, programming languages, and verification can help us better understand and constrain model behavior

This is a highly research-driven role focused on bringing rigorous systems thinking into frontier AI alignment work

You'll work on:

• Applying formal verification & program analysis ideas to model internals
• Building structured reasoning frameworks for interpretability research
• Exploring compiler-style abstractions for neural computations
• Developing tooling that combines interpretability with verification-inspired constraints
• Shaping an entirely new research direction inside the lab

Strong backgrounds include:

• Compilers or systems engineering
• Formal verification / theorem proving
• Security research or operating systems
• Research engineering with strong abstraction skills

PhD strongly preferred, but deep technical creativity matters most

This is not:

• Traditional enterprise verification work
• Pure theory with no implementation
• Large-scale ML infrastructure or training engineering

The environment is small, technical, and highly experimental, ideal for people who enjoy building research systems from scratch and applying rigorous reasoning tools to frontier AI problems