
1 minute read
Using Slack to Inform Your Gen AI
Overview
Salesforce’s Agentforce, along with ServiceNow, Google Vertex, and Amazon Bedrock, enables customer service automation through large language model (LLM) chatbots. These systems interpret tasks, execute predefined actions, and personalize responses using external data from databases and APIs However, managing and utilizing the dynamic knowledge within these systems remains challenging.
Context
At Arizona State University (ASU), employees share critical information in Slack channels, including data locations, policies, and troubleshooting tips. However, retrieving this knowledge via Slack’s keyword search is often inefficient, leading to lost information To address this, we developed an LLM chatbot system that helps users locate database tables, reports, and dashboards based on Slack discussions.
Purpose
This whitepaper demonstrates how Slack message data can power LLM chatbots to streamline knowledge management We highlight key insights, benefits, and drawbacks from this project to guide your enterprise data documentation and chatbot implementation projects.
Explore more in the full publication, “ASU Using Slack to Inform Your Gen AI,” available on Issuu.
Written by Varun Shourie, data scientist with the AI Acceleration team.