Applied Science Intern: Automated Reasoning for Cloud AI

Amazon; Boston (MA); On-site; USD 115,000 - 156,000; Full time

Benefits offered by this job

EAP

Mental Health Support

Medical Advice Line

401(k) matching

Job summary

Amazon is seeking an Applied Science Intern in Automated Reasoning and Formal Verification to contribute to the development of formal methods and automated reasoning techniques in a guided research setting. You will work on distributed proof search, SAT/SMT solving, and program analysis to verify cloud infrastructure and AI systems.

The role involves collaborating with scientists, applying AI-assisted approaches, and communicating research findings to diverse audiences.

Qualifications

• Enrolled in a PhD program
• Willing to relocate to internship location
• Commit to at least 12 weeks and 40 hours/week
• Proficient in Python; knowledge of C++, Haskell or Lean preferred
• Experience in mathematical logic, formal verification, SAT/SMT, theorem proving, model checking, or program analysis
• Demonstrated depth in a research area via publications or thesis

Responsibilities

• Design and implement algorithms and formal methods for automated reasoning
• Apply generative AI and machine learning techniques to enhance reasoning and verification
• Contribute to research on formal guarantees for AI systems and safety
• Publish findings at peer-reviewed conferences and journals
• Review outputs from AI tools and validate results

Skills

Python

C++

Haskell

Lean

Formal methods

SAT/SMT

Theorem proving

Model checking

Program analysis

Education

PhD candidate

Tools

Lean

Dafny

Isabelle

Rocq

Z3

SAT/SMT solvers

Job description

Amazon is seeking an Applied Science Intern in Automated Reasoning and Formal Verification to contribute to the development of formal methods and automated reasoning techniques in a guided research setting. You will work on distributed proof search, SAT/SMT solving, and program analysis to verify cloud infrastructure and AI systems.

The role involves collaborating with scientists, applying AI-assisted approaches, and communicating research findings to diverse audiences.