Job Description & Details
This is a heavy-hitting Applied AI Engineer role focused on building production-grade generative AI, agentic systems, and MLOps pipelines specifically for manufacturing, supply chain, and consumer goods. You aren't just building chat wrappers here; you're architecting deterministic-meets-probabilistic systems that integrate deeply with enterprise tech stacks like ERPs and PostgreSQL.
What You'll Actually Be Doing
You'll spend your days moving complex AI solutions out of fragile proof-of-concept notebooks and into secure, scalable production environments on Azure. Your main focus will be designing multi-step agentic workflows, building robust RAG pipelines with strict enterprise knowledge retrieval, and ensuring systems are continuously monitored for hallucination, drift, cost, and latency. Expect to interface with supply chain data, ERP systems, and business stakeholders to turn messy operational problems into reliable automated pipelines.
The Core Tech Stack
Python is your daily driver, backed by frameworks like PyTorch, TensorFlow, and scikit-learn for traditional ML alongside advanced LLM orchestration tools. You need deep, hands-on muscle memory with Azure OpenAI, Azure Database for PostgreSQL, vector databases, and containerized cloud-native deployments. The team is explicitly looking for engineers who can handle the messy reality of enterprise integration—connecting secure APIs, setting up event-driven architectures, and implementing rigid guardrails and tool-calling for AI agents.
Interview Expectations
The hiring manager is going to test your ability to balance probabilistic outputs with deterministic enterprise constraints. They will likely ask you to design a multi-agent system that interacts with an external inventory database while preventing hallucination and managing token costs and latency under load. You'll also need to defend your choices around evaluation frameworks and how you handle fallback logic when an LLM fails a guardrail check in production.
Application Advice
Make sure your resume doesn't just list buzzwords like 'GenAI' or 'RAG'—you need to explicitly showcase the lifecycle of models you've built, highlighting how you took them from experimental POCs to production monitoring with tools like Azure Monitor or dedicated LLMOps platforms. Emphasize any background in manufacturing, supply chain, ERP integrations, or consumer goods, as domain experience with supply chain data is a massive differentiator for this team.