Practical guides
How to ask AI questions about your business data
To ask AI about orders, customers and work history, first prepare accessible information that can be checked. Start the project with a business question, before choosing a tool.
Start with one question
Choose a recurring request: what happened to a vehicle, who asked you to return later, or which services customers reorder. Define who uses the answer and what they do next.
Connect the necessary sources
Repair history may require photos of job sheets. Sales analysis needs exports and spreadsheets; earlier agreements may be in an agreed email archive. FSS checks record quality, removes duplicates and connects information where possible.
Keep answers verifiable
An answer should lead back to a document or record. Missing or conflicting information needs to be visible. Unclear handwriting, missing fields and incorrect dates require review instead of a confident guess.
Test a limited process
Begin with an agreed sample and actual staff questions. Compare answers with the sources and the current way of working. Before expanding, define access, updates, corrections and deletion.
Choose the appropriate outcome
A one-off decision may need an analysis and report. Repeated questions may benefit from search and AI conversations; repeated data entry may need automation. FSS helps choose the scope for your task without requiring every capability at once.
What to bring to a discussion
- A question that currently takes too long to answer
- One document or anonymised export sample
- Who needs access and who will check answers
