Principal Platform AI Engineer, Healthcare Agentic AI

Required Skills & Experience

Top Requirements:

• 12+ years of software engineering at senior, staff or principal level, including 3+ years building production generative AI, RAG, agentic AI or AI platforms (not just POCs)
• Strong hands-on Python; C#/.NET preferred
• Agent orchestration, tool calling, context management, retrieval and evaluation
• Has set up or run spec-driven / contract-driven engineering practices
• Uses AI coding tools (Claude Code, GitHub Copilot) to speed up delivery while keeping controls in place
• Azure, Kubernetes/AKS, containers, APIs, microservices, event-driven architecture, CI/CD, observability
• Regulated-industry experience with sensitive data
• Has worked directly with vendor/SI engineers and validated architecture through code

Nice to Have Skills & Experience

Additional Skills:

• Healthcare: HIPAA, PHI, EMR, Provider/Pharmacy, FHIR/HL7
• MCP servers, clients and tools
• Microsoft Agent Framework, Semantic Kernel, LangGraph, AutoGen or LangChain
• Azure AI Foundry, Azure OpenAI, Databricks, vector databases
• LLMOps, AI evaluation, AI observability and AI FinOps
• Has built agent SDKs or reusable AI platform services

Job Description

Daily Responsibilities: A large healthcare organization is looking for a Principal Platform AI Engineer to lead hands-on engineering for its enterprise Healthcare Agentic AI Platform. This person sets up a spec-driven development model: architecture, APIs, data contracts, agent behavior and security controls are defined as specifications first, then turned into production code quickly using AI coding agents and automated quality gates. They design, code, test and productionize reusable platform capabilities on Azure, including agents, orchestration, MCP services, RAG, SDKs and evaluation, for use across multiple healthcare product lines. They work side by side with vendor engineers, review and contribute to their code, and make sure the organization keeps full code-level ownership of the platform. Everything is built for a regulated healthcare environment, with HIPAA, PHI protection, auditability and clear authorization boundaries for AI agents.

• Drive spec-driven engineering: Define specs, templates, standards and quality gates, and use AI coding agents to speed up spec-to-code, testing and review.
• Build the platform: Develop SDKs, APIs, runtime services, agent frameworks, MCP services and reference implementations.
• Engineer agents: Build production agents and multi-agent orchestration with tool calling, context management and human-in-the-loop.
• Deliver RAG and model integration: Build grounding, vector search and governed model access, and optimize prompts, latency and token cost.
• Run on Azure: Use AI Foundry, OpenAI, AKS, APIM, AI Search, Service Bus, Event Hubs and Key Vault with secure identity and private networking.
• Evaluate and monitor: Automate AI evaluations, tracing and cost tracking, and set budgets, rate limits and kill switches.
• Own vendor code quality: Review vendor code, flag security risks and technical debt, and drive fixes.