Job Description & Details
This is a heavy-hitting enterprise AI role focused on moving proof-of-concept models into secure, production-grade systems within regulated spaces like healthcare. If you are tired of just writing basic chat prompts and want to build actual agentic workflows and robust RAG pipelines that scale, this gig offers the technical depth you are looking for.
What You'll Actually Be Doing
Your day-to-day will revolve around architecting AI-powered services and integrating Large Language Models into existing enterprise environments. You will spend a good chunk of your time building and refining data pipelines, setting up vector search, and establishing retrieval-augmented generation patterns that pull securely from sources like SharePoint and Microsoft 365. It's not just about getting the model to work in a Jupyter notebook; you'll be dealing with the unglamorous side of engineering—setting up LLMOps, handling CI/CD for AI, monitoring drift, and collaborating tightly with security and compliance teams to ensure your agents aren't leaking enterprise data.
The stack here leans heavily into the Microsoft ecosystem. You must be comfortable working with Azure AI Services and Azure AI Foundry alongside standard Python frameworks like FastAPI, LangChain, and Semantic Kernel. On top of that, your data engineering chops need to be sharp because clean embeddings, API integrations, and robust vector database management are the backbone of everything you will build here.
Interview Expectations
The hiring manager will likely grill you on how you handle failure states in RAG systems, asking something like: 'How do you debug a hallucinating agent in production where the retrieval step pulled irrelevant context?' They are looking to see if you understand chunking strategies, reranking, and semantic caching, rather than just relying on out-of-the-box framework defaults. Another favorite will be around security: 'Walk me through how you would implement guardrails and RBAC when integrating an LLM with sensitive enterprise data sources like Confluence and SharePoint.' They want a pragmatic engineer who balances rapid AI innovation with strict compliance and data governance.
Application Advice
To get past the ATS, your resume needs to explicitly highlight terms like 'Retrieval-Augmented Generation (RAG)', 'Vector Embeddings', 'Azure AI', and 'LLMOps'. Don't just list technologies; write punchy bullet points detailing how you moved AI models from experimentation into secure production environments. If you have experience dealing with compliance frameworks or regulated industries like healthcare, make that front and center to prove you can navigate the enterprise red tape.