Copilot Managing Maintenance Intake by Zig
Short demo by Zig of a Copilot solution for maintenance intake: from a real estate unit looked up by address, a user starts the intake with an image, text, or both, and Copilot analyzes it and proposes tasks. The user reviews the explanation for the chosen task and confirms it, which creates the maintenance request. The attachments then hold the source image and the AI's explanation of what it sees.
- Problem statement and copilot solution overview
- Copilot capabilities for maintenance intake
- Starting maintenance intake with image
- Copilot analysis and task generation
- Viewing results and attachments
Overview
The video presents a Copilot solution for maintenance intake, motivated by a tenant reporting an issue in a house. Intake can start from natural language, an image, or a combination of both.
In the demo, the intake starts with an image. Copilot analyzes it, works out the problem, and generates tasks and possible solutions in a maintenance request. In the request’s attachments, the image is stored together with an explanation of what the AI sees in it. The video is about one minute long.
Key points
- The scenario is a tenant reporting an issue around the house. The old wizard-based intake was described as technical and knowledge-heavy.
- The Copilot solution does the intake from natural language, an image, or both, and proposes a possible solution for the fix.
- The intake starts from a real estate unit that the user looks up by address.
- In the demo, an image of a broken door is used; after Generate, Copilot analyzes the image and suggests possible solutions.
- The user can check the explanation of why a task was chosen and confirm it, which creates the maintenance request.
- The request’s attachments show the source image and an AI explanation of what it sees in the image.
Features
Every feature below: status not stated, demoed.
| Feature | At |
|---|---|
| Copilot-powered maintenance intake | 0:13 |
| Image-based maintenance issue analysis | 0:35 |
| Maintenance request attachments and AI explanation | 0:46 |
Quotes
- 0:01 “for the reporting of an issue in around the house by a tenant. Uh we have an intake or sick operations that is done”
- 0:13 “What we’ve created is a co-pilot solution that can do the intake based on natural language, an image or a combination both to uptake”
- 0:35 “copilot can look at the image, analyze it, and see what possible uh solutions there are.”
- 0:46 “We can check for the explanation why it has chosen that uh task and we can confirm it. This will create the maintenance request.”
- 0:57 “see a little bit of an explanation what the AI is seeing on the image to give a little bit of a of a”
Presenters (as heard): Zig.