Description

You will build and maintain autonomous AI coding agents and integrate them into the production workflow.

This role is on-site.

Responsibilities

• Develop and deploy AI agents that execute multi-step coding tasks, including writing code, running tests, fixing errors, and submitting pull requests.
• Implement and manage Model Context Protocol (MCP) servers to connect AI tools with external data sources and codebases.
• Write effective prompts for code generation, defining constraints, context files, and output formats.
• Identify and mitigate failure modes in AI-generated code, including security risks like prompt injection, SQL injection, and hardcoded secrets.
• Establish evaluation frameworks for automated testing of AI outputs to ensure production readiness.

Required Skills

• 5+ years of software engineering experience with a Bachelor's degree in Computer Science or equivalent practical experience.
• Daily use of at least two AI coding tools (e.g., GitHub Copilot, Cursor, Claude Code, Windsurf, Aider) in production for the last 12 months.
• Hands-on experience with AI-driven code review tools such as GitHub Copilot Code Review, CodeRabbit, Greptile, or Qodo.
• Proficiency in SQL and backend development.
• Experience using GitHub Copilot as a primary development tool.
• Understanding of AI agent architectures (e.g., Devin, OpenHands, Cursor Agent mode) and their operational limits.
• Ability to describe specific security, performance, and architectural failure modes of AI-generated code.

Preferred Skills

• Experience building or configuring MCP servers.
• Familiarity with evaluation frameworks for AI output quality and safety.

Key Skills

Education

Bachelor's degree