AU Class
AU Class
class - AU

Lessons Learned from Developing AI Chatbot for Multibillion-Dollar Semiconductor Fabrication Plants

Share this class
Search for keywords in videos, presentation slides and handouts:

    Description

    Imagine a mega project that generates nearly 2,000 Revit files in total, and some getting updated in daily basis. The speed and volume of data produced in multi-billion semiconductor fabrication plant project are staggering. Extracting valuable insights from this vast data pool must be effortless and straightforward, enabling owners to communicate effectively with designers and contractors to ensure project success. Samsung Electronics envisioned an AI Chatbot capable of understanding and interpreting project information as the ultimate solution. In this session, we will introduce our journey of developing an AI Chatbot using Autodesk Platform Services (APS) and Amazon Bedrock. Discover the challenges we encountered in implementing multi-agent collaboration for Text-to-SQL and effectively ingesting large amounts of data to feed the Large Language Model (LLM). Additionally, learn how we adapted our approach to make the chatbot truly beneficial for the project.

    Key Learnings

    • Learn key challenges encountered during the development of AI Chatbot for megaproject.
    • Learn insights gained from developing AI Chatbot using Amazon Bedrock and APS.
    • Learn insights gained from developing AI agents and multi-agent orchestration system for Text-2-SQL.
    • Learn insights on how AI chatbot system can benefit megaproject.

    Downloads