Job Description & Details
This is a hands-on AI/ML engineering role focused on building and scaling LLM-powered features and agentic workflows for enterprise applications. If you enjoy bridging the gap between raw machine learning research and rock-solid production code, this gig cuts straight through the noise.
What You'll Actually Be Doing
You will spend your days designing RAG pipelines, optimizing prompt chains, and making sure AI-generated outputs don't hallucinate in front of enterprise clients. You won't just be throwing Jupyter notebooks over the fence; you'll be writing scalable Python backends, integrating APIs, occasionally tweaking React frontends, and dealing with the unglamorous reality of latency, cost-efficiency, and MLOps monitoring in production.
The Core Tech Stack
You need absolute fluency in Python and production-grade API development, coupled with deep, practical experience building LLM applications using vector databases and RAG frameworks. Familiarity with cloud infrastructure like AWS, Azure, or GCP is non-negotiable because you're expected to deploy workloads that don't crash under load. Knowing your way around React and MLOps tooling gives you a massive edge here.
Interview Expectations
Expect to be grilled on how you handle retrieval accuracy and context window limits in RAG pipelines, so be ready to explain how you evaluate semantic search quality beyond simple keyword matching. The hiring manager will likely dig into a past production failure where an AI service spiked in latency or cost, and they'll want to hear how you refactored the architecture, implemented caching, or optimized token usage to fix it.
Application Advice
Skip the generic objective statement and make sure your resume highlights specific metrics around systems you've taken to production, such as handling X requests per second or reducing LLM latency by Y percent. Ensure keywords like Retrieval-Augmented Generation, agentic workflows, Python, MLOps, and scalable APIs are explicitly called out to clear the initial ATS filters.