- Company Name
- GitHub
- Job Title
- Software Engineer III, Copilot Agents
- Job Description
-
Job Title: Software Engineer III, Copilot Agents
Role Summary: Design, develop, and maintain advanced AI‑powered code analysis and remediation engines, including CodeQL, Copilot Code Review, and Copilot Autofix. Drive the integration of large language models (LLMs) with static analysis to detect security and quality issues early, and to generate actionable fix suggestions for developers in a distributed, remote-first environment.
Expectations: • Deliver high‑quality, production‑ready code for multi‑language support (C/C++, C#, Go, Java, JavaScript/TypeScript, Kotlin, Python, Ruby, Swift). • Maintain and evolve CodeQL query language implementations and source code extractors. • Build, test, and refine LLM‑based detection and fix agents, ensuring low false‑positive rates and robust performance. • Collaborate cross‑functionally with product managers, designers, and technical writers to shape product direction and user experience. • Publish findings, perform experiments, and document best practices for internal stakeholders and open‑source community.
Key Responsibilities: • Maintain and extend source code extractors for supported languages, staying current with language evolutions. • Write and optimize CodeQL queries and libraries to detect vulnerabilities and undesirable patterns. • Ensure quality of LLM‑generated fix suggestions and review comments, iterating on model integration and prompt design. • Experiment with and evaluate LLM‑powered detection engines against traditional static analysis, reporting results. • Design and implement LLM‑based code review and fix generation agents for Copilot Code Review and Autofix. • Engage with internal users and customers to gather feedback, run demos, and validate feature impact. • Contribute to documentation and internal knowledge bases around AI‑powered code analysis.
Required Skills: • Strong software engineering experience (5+ years) working with large codebases and static analysis tools. • Proficiency in at least one target language (e.g., Java, Python, JavaScript). • Experience with building static analysis or query languages (e.g., CodeQL, Clang, ESLint). • Familiarity with large language models, NLP pipelines, and prompt engineering for code. • Experience in building production‑grade ML/AI services, including model deployment, monitoring, and evaluation. • Ability to write clear, maintainable code and documentation; proficiency with Git workflows. • Excellent problem‑solving, debugging, and communication skills.
Required Education & Certifications: • Bachelor’s (or higher) degree in Computer Science, Software Engineering, or related field. • Prior experience with LLMs, static analysis tooling, or AI‑based code assistance is a strong plus.