Software Engineer - AI (Level determined by experience)

Synack is hiring a Software Engineer - AI to design and ship autonomous AI agents that conduct penetration testing in sandboxed environments. The role owns systems from development through production, improves agent reliability and efficiency, and partners with security researchers to translate their expertise into agent capabilities.

Responsibilities:

• Build AI agents that automate penetration testing: planning, tool use, reasoning over findings, and acting safely within scope
• Stand up and run the sandboxed cyber-lab and range infrastructure, containerised on Kubernetes, where these agents execute offensive tooling
• Ship features from ideation through production, including evals, guardrails and monitoring
• Improve how our agents work: prompting, tool orchestration, evaluation, and cost
• Lead code reviews and design sessions, and help the engineers around you grow
• Work closely with our security researchers to translate how they hack into how our agents operate

Requirements:

• 7+ years of experience in software engineering
• Proven ability to build and scale production systems
• Strong Python, polyglot codebase
• Fluent or willing to learn Go
• Experience building LLMs an agentic frameworks leveraging tools like RAG, MCP- from dev environment to production
• Demonstrate ability to create reliable agents (prompting, evaluation, guardrails, understanding of token and cost optimization
• Capable leveraging Docker, Kubernetes, isolated sandbox, cloud experience GCP preferred
• Brings a disciplined testing approach to microservices gPRC, API design, async messaging (Pub/Sub, Kafka)
• Passionate desire to understand how systems get compromised
• Clear communication
• Collaborative ability to partner across teams
• We can only hire U.S. citizens for this position due to federal government contract requirements
• Experience in a fast-paced startup environment
• Experience in Cyber Security