Job Description & Details
This is a hands-on, senior-level AI engineering role where you aren't just writing Jupyter notebooks; you are taking modern generative and classical machine learning models all the way to production. If you enjoy owning the full stack from raw data ingestion and RAG architecture to Azure deployments and cost-latency observability, this gig gives you the autonomy to build real, trusted AI systems.
What You'll Actually Be Doing
Expect to spend your days knee-deep in architecting retrieval-augmented generation pipelines, optimizing LLM agentic workflows, and setting up rigorous evaluation harnesses. You'll juggle prompt engineering, fine-tuning, and vector store tuning alongside traditional MLOps tasks like drift detection and CI/CD automation. Half your challenge will be writing clean code and using AI-assisted tools like Claude Code and Cursor effectively, while the other half involves talking to stakeholders to scope problems and make sure your AI solutions actually solve tangible business bottlenecks.
The Core Tech Stack
You'll need a rock-solid foundation in Python, PyTorch, and core data libraries, paired with heavy production experience using major LLM providers like Anthropic Claude or Azure OpenAI. Because everything lives in the cloud, deep familiarity with Azure, Docker, Kubernetes, and vector databases like Azure AI Search or pgvector is non-negotiable. They also expect you to know your way around GitHub Actions or Azure DevOps to keep automated testing and deployment humming.
Interview Expectations
When you get on the technical rounds, be ready to defend your architectural choices around RAG chunking strategies and latency optimization when scaling vector search. They'll likely ask you how you implement guardrails and handle prompt injection or PII in regulated environments, so have a concrete war story ready about a time you secured a production LLM endpoint. The hiring managers are secretly testing whether you rely too blindly on AI code assistants or if you possess the deep underlying fundamentals to debug complex distributed failures yourself.
Application Advice
Your resume needs to skip the generic buzzwords and laser-focus on production outcomes. Explicitly highlight numbers around the ML and AI systems you've shipped—mention the vector stores you've scaled, the specific Azure services you've deployed to, and how you measured model drift or cost reduction. Make sure keywords like RAG, Azure AI Search, PyTorch, Docker, and CI/CD are prominently featured to easily clear the ATS filters before a senior engineer reviews your background.