LLMs are not deterministic. They do not always return the same answer—or even the right one. But they are still powerful for industrial use cases. To unlock their potential, you need hands-on experience in a space where you can experiment, iterate, and learn with your peers. That is what this workshop is all about.
Join Aron Semle, HighByte CTO, for a 3-hour hands-on workshop where you'll build and run your own AI agent, fully integrated with an MCP server and running locally on your machine. By the end of the workshop, you will have a working setup you can continue to experiment with—and a solid understanding of how LLMs, agents, and MCP interact with one another.
This workshop is recommended for manufacturing IT professionals, data engineers, and system architects at industrial companies.
What Is Included
- Access to a 2-hour trial of HighByte Intelligence Hub
- GitHub project with everything you need to get started
- Advance instructions for installing the required software
- Access to the recording to watch anytime on-demand
Minimum System Requirements
- Java SE 21 or OpenJDK 21
- HTTP server for hosting the Intelligence Hub frontend
- 4 GHz processor
- 1 GB RAM
- 4 GB available disk space
- Network capable (TCP)
- Paid API key for an LLM (e.g., OpenAI, Anthropic, Google Gemini) OR a machine that can run a small language model (i.e., 4B parameter)
Agenda
- Review the fundamentals of MCP, function calling, and the agentic loop
- Set up HighByte Intelligence Hub at the MCP Server for your agentic project
- Discuss the fundamentals of MCP tools, test these tools using MCP Inspector, and prepare advanced tools for the Agent
- Set up a Small Language Model (SLM) to explore how it interacts with and calls tools
- Review best practices for designing open, vendor-agnostic architectures
- Get tips on how to move from proof-of-concept to enterprise implementation
