Project Summary

Led the design and delivery of a domain-specific GenAI application using a microagent-based RAG architecture to support industrial engineers and operators. The project successfully completed two pilot phases and is now being transitioned to a global product team.

Technologies Used

Microagent architecture, Retrieval-Augmented Generation (RAG), LangGraph, LlamaIndex, Large Language Models (LLMs), observability frameworks for RAG systems, heuristic memory control for LLMs.

Industry Applied

Industrial automation and manufacturing (e.g., smart factories, process engineering). Also applicable to energy, utilities, and heavy machinery sectors.

Benefits Delivered

• Reduced engineering workload processing time from hours to under 30 seconds
• Enabled real-time contextual assistance for operators
• Improved knowledge reuse across global teams
• Enhanced onboarding for new engineers through instant access to validated operational knowledge