data
Posted 22 hours agoAI Engineer
at Metriport
San Francisco, United StatesOn-site
Responsibilities
- - Build internal agents and copilots that take work off the team's plate: monitoring the platform and preparing fixes, triaging customer issues, drafting responses, generating compliance evidence for compliance audits, summarizing PR activity, on-call assistance — anything where a human is currently doing rote pattern-matching.
- - Design the eval, observability, and guardrail layer for everything you ship.
Requirements
- AI ENGINEER Metriport is an open-source data intelligence platform that helps healthcare organizations access and exchange patient data in real-time.
- ABOUT THE ROLE This is the first AI Engineer role at Metriport, and one of the highest-leverage hires we'll make this year.
- We're looking for an engineer who treats AI agents as leverage — someone who will literally change how we ship products and how we operate.
- - Collaborate with the rest of the engineering team to integrate AI to our products in ways we haven’t even conceived yet.
- AI systems that touch production code, customer data, or money need to be measurable, debuggable, and safe to roll back. - Work directly with founders and team leads to figure out which workflows are highest-leverage to work on next, and ruthlessly prioritize.
- You know what fails in production, and you've got opinions about evals, prompt versioning, tool use, retries, and cost control. - You're a strong full-stack engineer first, AI engineer second.
- TypeScript / Node.js, Python, AWS, and SQL aren't going to scare you. - You think in workflows and systems.
- You can sit in on a customer support call or a PR review and immediately spot the three steps a model could be doing. - You have a strong sense of ownership.
- experience with healthcare data, compliance environments (SOC 2, HIPAA, HITRUST), or shipping AI-enabled workflows at an early-stage startup. - You’re located in San Francisco or the Bay Area (or willing to relocate).
- OUR TECH On the frontend, we use React - on the backend, we rely on Node.js and TypeScript for writing core business logic.
- We deploy a wide range of AWS cloud services (ie ECS, Fargate, Lambda, etc), and manage our infrastructure as code with AWS CDK.
- Data lives in PostgreSQL, DynamoDB, S3, Snowflake, FHIR server, and more.