• Own your work end-to-end. You work closely with the product manager, who owns the roadmap and defines what’s worth solving. Your job is to take that direction, talk to users and customers, write your own spec or requirements (often with the help of an AI agent), build it, and ship it — without waiting for every detail to be handed to you.
• Talk to users — and do it well. Customer SMEs, clinical operators, payer engineering teams, integration partners. You ask the right questions, hear what’s actually being said, and translate it into a clear engineering problem.
• Make UX and design calls on small-to-medium decisions independently. Partner with design on larger ones.
• Instrument the work you ship. Adoption, latency, error rates, time-to-value — you define what “working” means for your feature and make it measurable. Success metrics are owned by the PM; your job is to make sure the data is there to evaluate them.
• Build full-stack: services and APIs in C# / ASP.NET Core, data in SQL Server and Cosmos DB, async messaging via Azure Service Bus, deployed on Azure. Front-end work as needed; the product surface matters as much as the back-end.
• Ship with AI tools as primary accelerators. Cursor, Claude Code, GitHub Copilot — daily use, not novelty. Move from idea to working prototype in hours.
• Build LLM-powered features into the product where they solve a user problem better than deterministic code: RAG over clinical content, agentic patterns, prompt engineering for production. Define guardrails for PHI, auditability, and clinical determinism.
• Hold the engineering bar. Your code goes through the same review, security, and reliability standards as any other engineer at MCG.

What We're Looking For:

• 3+ years of professional software engineering experience. (Years are a floor, not the primary signal — a 3-year engineer who has shipped end-to-end production features with real user impact is more relevant than a 7-year engineer who hasn’t.)
• Track record of at least one feature owned end-to-end — idea, ship, and measured impact. You can describe what you built, what you learned, and what you’d do differently.
• Strong full-stack engineering depth in C# / ASP.NET Core (MCG’s primary stack), or comparable modern OO language Vue.js or equivalent modern front-end framework (React, Angular, Svelte)
• Microservices, distributed systems, async messaging
• SQL Server or equivalent relational database (PostgreSQL, Oracle, MySQL)
• Cosmos DB or equivalent document/NoSQL store (MongoDB, DynamoDB, Firestore)
• Azure Service Bus or equivalent messaging platform (RabbitMQ, Kafka, AWS SQS/SNS)
• Cloud-native development on Azure or comparable cloud (AWS, GCP)
• API design (REST), OAuth 2.0 / OIDC
• Effective at communicating with non-engineers customers, clinical operators, support, PMs.
• Progress over movement. You instinctively distinguish work that advances outcomes from work that just feels productive
• Strong written communication. You document decisions and product context for the team.

• Evals discipline — offline golden sets, LLM-as-judge, regression detection, online evaluation patterns. You can answer “how do you know your LLM feature is getting better or worse over time?” with specifics.
• Agent reliability in production — tool error handling, state management across long-running agent loops, retry/fallback design, knowing when to break out of an agent loop rather than spin indefinitely.