LLM Mesh agentic applications#

This tutorial series demonstrates how to build agentic applications using the LLM Mesh in Dataiku. Through an example of building a customer information assistant, you’ll explore three parts of an agentic workflow:

  1. Tools - How to define and use tools with the LLM Mesh

  2. Agents - How to create an LLM-based agent that uses multiple tools

  3. Webapps - How to build a web application with the agent

Prerequisites#

  • Dataiku >= 13.2

  • LLM Mesh connection to a provider that supports tool calls (tested with OpenAI GPT-4; you can find a compatible list here)

  • Project permissions for “Read project content” and “Write project content”

  • An SQL dataset pro_customers_sql (a CSV file can be downloaded here containing customer data

  • Python environment with the following packages

    dash # tested with 2.18.2
dash-bootstrap-components # tested with 1.6.0
duckduckgo_search # tested with 7.1.1