Applied Intelligence Engineer
Part-time: Up to 20 hours per week Remote | $80–$100 per hour
An opportunity is open for an Applied Intelligence Engineer to contribute to a cutting-edge AI lab focused on automated code understanding and large-scale software reasoning. The role supports a high-impact code review project and is well-suited for engineers with deep experience in navigating and analyzing massive software repositories across multiple languages.
A leading AI research team is hiring for this role to design, evaluate, and improve intelligent systems that reason about real-world, production-scale codebases. The work involves deep code analysis, agent task design, and close collaboration with research engineers.
Key Responsibilities
- Analyze and document the structure of large codebases, including modular components, dependency graphs, and architecture patterns
- Design evaluation tasks for AI agents that simulate real-world engineering scenarios, such as:
- Navigating project structures and build systems
- Refactoring or documenting code
- Detecting anti-patterns or bugs
- Performing code reviews
- Collaborate with researchers to define benchmarks and success criteria
- Offer language-specific insights (Java, Python, C++, TypeScript, Kotlin preferred)
- Review agent outputs and provide detailed feedback on behavior, limitations, and edge cases
Qualifications
- Extensive experience working in production-scale codebases, such as PyTorch, Kubernetes, or similar
- Strong fluency in at least one major language (Java, Python, TypeScript, C++, Kotlin)
- Familiarity with complex code structures, modular repositories, and CI/CD workflows
- High-level understanding of software quality, maintainability, and architecture
- Experience with code reviews, static analysis, or intelligent dev tooling is a plus
- Ability to work independently and communicate clearly in writing
Work Setup
- Part-time: Up to 20 hours per week
- Fully remote and asynchronous; flexible schedule
- Contract-based with weekly payment through Stripe Connect
- Initial term: Project-based, with possible extensions