Job Description & Details
This is a serious, heavy-duty engineering role focused on building production-grade LLM applications, RAG pipelines, and agentic workflows rather than just tinkering with prompt engineering. If you enjoy wrestling with distributed systems, complex document intelligence, and keeping AI models hallucination-free in production, this gig will keep you on your toes.
What You'll Actually Be Doing
You will spend your days designing and deploying LLM-powered applications, stitching together multi-agent orchestrations, and fine-tuning retrieval-augmented generation (RAG) pipelines. Expect to spend a lot of time optimizing chunking strategies, embeddings, re-ranking, and grounding mechanisms to ensure clinical and operational documents are parsed with extreme accuracy. You will also build out human-in-the-loop validation loops, instrument ML observability dashboards to track token costs and latency, and containerize Python microservices to run smoothly on Kubernetes.
The Core Tech Stack
You need to be completely fluent in Python, containerization tools like Docker and Kubernetes, and modern cloud ecosystems like Azure AI or AWS SageMaker. Hands-on experience building RAG architectures utilizing vector databases, knowledge graphs, and agentic frameworks is non-negotiable here. You will also need a solid grip on data engineering platforms like Spark or Databricks, alongside CI/CD tooling like GitHub Actions, because deploying these models reliably is just as important as writing the core algorithms.
Interview Expectations
Expect the hiring team to grill you on your approach to scaling RAG pipelines and handling context window limitations. They will likely ask how you manage document intelligence workflows when dealing with messy, unstructured clinical text and how you architect fail-safes using human-in-the-loop review gates. The interviewer isn't just looking for textbook definitions; they want to hear war stories about how you've debugged semantic drift, optimized token usage, or handled hallucination mitigation in a live production environment.
Application Advice
Make sure your resume doesn't just list every AI framework you've ever heard of—focus strictly on systems you've actually built and deployed to production. Explicitly highlight keywords like RAG, vector databases, multi-agent orchestration, Python, and ML observability. If you have experience with responsible AI, bias mitigation, or processing complex documents at scale, put those front and center to immediately clear the ATS and catch the hiring manager's eye.