Note: The job is a remote job and is open to candidates in USA. United Network for Organ Sharing (UNOS) is hiring a Senior AI/ML Engineer to build and operate the Azure infrastructure, development patterns, and production agents that support AI across the organization. The role focuses on deploying secure AI and machine learning systems, establishing evaluation and monitoring practices, and enabling engineering teams to develop and ship AI solutions.
Responsibilities
- Provision AI infrastructure on Azure as code, including model deployments in Microsoft Foundry and Databricks, GPU capacity, agent runtimes, as well as the networking, identity, and secrets around them
- Stand up open-weight and frontier models on UNOS infrastructure under a pattern approved by Security and Privacy for protected health information (PHI)
- Develop the technical policies and build controls into the platform itself: per-agent identities, least-privilege access, private endpoints, allow-listed models, and audit evidence produced as a byproduct of normal use
- Run shared model access across providers, with routing, quotas, tracing, and cost attributed to each team and use case
- Build sandboxed environments where engineers run coding agents against real repositories, with no production credentials inside
- Define UNOS’s standard patterns for agentic development and publish them as working templates and reference implementations
- Integrate AI into CI/CD for pull-request review, test-failure triage, and security findings
- Build and ship UNOS’s own agents, in UNOS One products and for internal operations, including tool access, retrieval, orchestration, and human approval where a decision requires it
- Build the path from prototype to production in UNOS One, so a working proof of concept ships without being rebuilt
- Own evaluation and release for agents and conventional models alike, including test sets, staged rollout, rollback, and monitoring
- Baseline and track the time from idea to working proof of concept, and the quality of what ships
- Teach agentic development across UNOS through pairing, working sessions, and code review, and mentoring engineers in the AI function
- Recommend build-or-buy decisions on AI platforms and tools, with cost and risk stated
- Perform other duties and support additional administrative priorities as assigned, consistent with the general scope and level of the position
Skills
- Approximately six (6) years of experience building and running production systems, including at least three (3) years owning cloud infrastructure as code
- Production cloud depth, Azure preferred (equivalent AWS or GCP experience considered), with Terraform or Bicep: virtual networks, private endpoints, managed identities, role-based access, secrets, and policy
- Strong handle on fundamental principles of AI/ML (e.g. train/test/validation, overfitting, calibration, eval methods)
- Hands-on provisioning of AI workloads, including model serving endpoints, quota and capacity, GPU compute, and cost controls
- An LLM-based system built and run in production, past the demo stage, including tool calling or agents, retrieval, evaluation, and tracing
- Experience working under HIPAA, HITRUST, SOC 2, FedRAMP, or a similar control framework, translating controls into configuration and evidence
- Daily practice with coding agents such as Claude Code, GitHub Copilot, Codex, or Cursor
- Strong proficiency in Python, containers, and Kubernetes or an equivalent orchestrator
- Demonstrated record of changing how a team works without managing it, through establishing useful repeatable patterns that reduce team burden and shorten time to solution or insight
- Ability to choose what to build first with incomplete information and ship working versions early
- Experience with Microsoft Foundry, Foundry Agent Service, Azure OpenAI, or Azure API Management as an AI gateway
- Experience with Databricks, including Unity Catalog, Model Serving, and MLflow
- Experience serving open-weight models with vLLM, SGLang, or similar, including quantization and GPU capacity planning
- Experience building MCP servers or other tool interfaces over existing APIs and legacy systems
- Experience with agent security, including prompt-injection defenses, tool permissions, and sandboxing
- Experience operating conventional machine learning: training pipelines, model registries, and drift monitoring
- Experience with HL7 FHIR or EHR integration
- Organ donation and transplant experience is a plus
- Bachelor's Degree in Computer Science,Engineering or a related field preferred
Company Overview