Job Description & Details
This is a heavy-hitting technical lead role focused on building real enterprise-grade Agentic AI and RAG architectures rather than just stitching together simple wrapper apps. You will be sitting right at the intersection of modern LLMs and complex enterprise ERP/Supply Chain systems, making it a crucial hire for modernizing legacy business logic.
What You'll Actually Be Doing
Your day-to-day will involve architecting multi-agent systems, writing robust Model Context Protocol (MCP) servers, and tying LLMs directly into Oracle Database 26ai features like Vector Search and Select AI. You aren't just writing Python scripts; you are designing secure data pipelines, optimizing PL/SQL APIs, and ensuring these AI agents can safely interact with core enterprise modules like Procurement, Inventory, and Manufacturing without breaking business workflows.
The Core Tech Stack
You absolutely must know your way around modern orchestration frameworks like LangChain or LangGraph, alongside deep experience in Retrieval Augmented Generation (RAG). Beyond the standard Python and LLM tooling, heavy proficiency in Oracle Database 26ai, Vector Search, and advanced SQL/PL/SQL is non-negotiable because this infrastructure relies on in-database AI capabilities and direct ERP integration.
Interview Expectations
Expect the hiring team to grill you on how you handle state management, error recovery, and infinite loops in multi-agent systems or complex LangGraph workflows. The interviewer is secretly looking to see if you understand the actual failure modes of LLMs in production environments and whether you can architect secure fallback mechanisms when an agent hallucinates a tool call against an ERP system.
Application Advice
Make sure your resume doesn't just list generic AI buzzwords, but specifically highlights your hands-on experience with RAG pipelines, database-level vector search, and enterprise integrations like ERP or Supply Chain platforms. Explicitly drop keywords like LangGraph, MCP, Oracle Database 26ai, and PL/SQL right into your professional summary so your application easily clears the initial ATS filter.