Job Description & Details
This is a hands-on engineering role focused on building production-grade Generative AI applications rather than just tinkering with prompts in a playground. You will be knee-deep in architecting RAG pipelines, deploying AI agents, and integrating heavy-duty data engineering stacks in the healthcare space. If you like bridging the gap between raw LLM APIs and enterprise data systems, this gig will keep you thoroughly engaged.
What You'll Actually Be Doing
Expect to spend your days designing document ingestion workflows, optimizing embeddings, and plumbing vector databases to ensure LLMs retrieve the right context without hallucinating. You will build and scale APIs for AI-powered features, hook up agentic frameworks like LangChain or AutoGen, and deal with the messy reality of data cleaning using PySpark and Databricks. A significant chunk of your effort will also go into setting up guardrails, monitoring LLM drift, and ensuring responsible AI practices within a heavily regulated domain.
The Core Tech Stack
Python is your daily driver here, paired with heavy data lifting via PySpark, Databricks, and SQL. You need real, battle-tested familiarity with major LLM providers like OpenAI, Azure OpenAI, or Anthropic Claude. On top of that, fluency with vector databases like Pinecone, Qdrant, or ChromaDB is non-negotiable since you'll be building scalable retrieval systems. If you've spent time wiring up agentic workflows with LangGraph or CrewAI and deploying on Azure, you'll feel right at home.
Interview Expectations
When you sit down with the engineering lead, expect them to grill you on how you handle latency and context window limits in production RAG systems. They might ask you to whiteboard a solution for optimizing token usage while maintaining high retrieval accuracy over millions of unstructured documents. They aren't looking for textbook definitions; they want to hear war stories about how you debugged a silent failure in an embedding pipeline or managed hallucinations in a multi-agent workflow.
Application Advice
Skip the generic AI buzzwords and tailor your resume to highlight the heavy data engineering lifting behind your AI projects. Make sure explicit terms like PySpark, Databricks, vector database names you've used, and specific orchestration frameworks like LangChain or LangGraph stand out immediately so you clear the ATS filters. If you have any background dealing with healthcare data compliance, claims, or clinical workflows, push that to the top—it's the secret sauce they are looking for.