Build AI Agents for MSP Help Desk & Internal Knowledge Bots

Analyzing MSP tickets with GPT5

The host demonstrates feeding fictional MSP help desk tickets into GPT5 to identify trends, causes, and resolutions, aiming to create a summary for an AI agent.

Creating an AI agent from a system prompt

The video shows how to manually build a new AI agent in the HATZ Workshop by copying a system prompt generated from the ticket analysis, choosing an AI model, and naming the agent.

Testing the MSP help desk agent

A sample support ticket is fed to the newly created MSP help desk agent, which quickly provides triage steps, questions for the user, and next actions, demonstrating its efficiency.

Agent's conversational memory benefit

The host highlights the key advantage of an agent over a simple chat: its ability to retain conversational memory, allowing for follow-up questions and continued interaction within the same ticket context.

Internal finance knowledge bot

The video expands on the agent's utility, suggesting its use as an internal knowledge bot for finance teams, providing answers based on company processes and financial documents.

HR onboarding and company knowledge bot

Another use case is presented: an HR onboarding tool where new employees can ask questions about company policies, processes, and documentation, streamlining their integration.