Job Description & Details
This is a heavy-duty production engineering role focused on moving experimental LLM setups into scalable, secure AWS environments. If you are tired of just writing toy prompts and actually want to build robust RAG pipelines and modular Python services that handle real enterprise traffic, this gig is worth looking into. You will be bridging the gap between cutting-edge AI models and boring, reliable backend infrastructure.
What You'll Actually Be Doing
Expect to spend your days knee-deep in Python services, wiring up OpenAI APIs and orchestrating agentic workflows inside AWS. Your main challenge will be architecting reliable Retrieval-Augmented Generation (RAG) pipelines that don't choke on latency or hallucinate when connected to enterprise data stores. You will also be wrestling with IAM policies, setting up CloudWatch observability, and migrating messy custom GPT prototypes into bulletproof microservices running on ECS or EKS.
The Core Tech Stack
Python is non-negotiable here, and you need to know how to write clean, modular code rather than just data science notebooks. On the infrastructure side, AWS is your playground—specifically services like Lambda, S3, ECS, and OpenSearch for semantic search. Familiarity with LangChain, vector databases, and OpenAI's API integration patterns is critical because the team expects you to hit the ground running with modern LLM orchestration and RAG workflows.
Interview Expectations
When you jump on a call with the engineering team, expect them to grill you on how you handle failure modes in LLM pipelines, such as what happens when the OpenAI API goes down or latency spikes. They are going to ask you to design a secure, low-latency RAG system from scratch on AWS, and they will be listening closely to see if you actually understand VPC configurations, IAM least privilege, and vector search scaling, or if you just know how to call an API wrapper.
Application Advice
Skip the generic resume fluff and immediately highlight your hands-on experience deploying Python microservices that actually talk to AWS infrastructure and LLMs. Make sure keywords like RAG, OpenAI, AWS Lambda, ECS, and semantic search are baked right into your bullet points, because the ATS is definitely scanning for them. If you have migrated an experimental AI project into a production environment before, put that front and center.