Job Summary:
Vijil is an AI software startup on a mission to help organizations build and operate autonomous agents that they can trust. The role involves owning the backend platform services, ensuring reliability in production, and collaborating with a small team to deliver solutions to customers.
Responsibilities:
• Own production end-to-end: cloud infrastructure, deployments, observability, on-call.
• Own CI/CD: build, test, release, and image promotion across every repository we ship.
• Build, deploy, and maintain the AI agents and libraries that make up our product.
• Write production-quality backend code in Python.
• Set the engineering bar: clean architecture, real tests against real systems, no silent failures.
• Cut through cross-repo work that no one else has the context to finish.
Qualifications:
Required:
• 10+ years of professional software engineering, primarily backend.
• Core cloud development competency. You have designed, deployed, and operated production services on a major cloud. You understand networking, identity (IAM or equivalent), storage (S3 or equivalent), and compute (EC2/EKS or equivalent) at the level needed to debug them under pressure. This is non-negotiable.
• Hands-on experience building or operating agentic AI systems in production — LLM-backed agents that call tools, coordinate with other agents, and run as services. You can speak concretely about what failed in production and how you fixed it. This is non-negotiable.
• Strong Python, including modern async, type hints, and production frameworks such as FastAPI and Pydantic.
• Ownership of CI/CD pipelines you built or rebuilt yourself (GitHub Actions, AWS CodeBuild, or equivalent).
• Container-based deployment and orchestration at production depth: Docker, Kubernetes, and Helm.
• Working fluency with relational databases (PostgreSQL preferred), including writing migrations and debugging slow queries.
• Disciplined Git workflow, code review hygiene, and semantic commit history. GitHub link appreciated.
• Strong written communication. You write design docs that other engineers actually read.
Preferred:
• Familiarity with agent interoperability standards such as MCP (Model Context Protocol) or A2A (Agent-to-Agent).
• Experience with agent development frameworks such as Google ADK, LangGraph, or equivalent.
• Production observability experience with OpenTelemetry or equivalent.
• Experience operating a multi-service platform owned by a small team.
• Open-source maintainer or contributor history.
• Prior work on AI infrastructure, agent platforms, or evaluation systems.
• Comfort working directly with applied scientists and translating research into production code.
Company:
Vijil is an AI software startup that helps developers build and operate intelligent agents that people can trust. Founded in 2023, the company is headquartered in Menlo Park, USA, with a team of 11-50 employees. The company is currently Early Stage.