AI features
Brandow Storage includes AI workflows for both setup and daily data operations.
This guide explains what each feature does, when to use it, and how to get better results.
AI Table Creation
What it does
You describe a data structure in plain language. AI proposes a Table with columns and data types.
Best use cases
- New process setup (CRM, inventory, ticketing, project tracking)
- Quick prototyping before committing to a final structure
- Converting rough process ideas into a concrete data model
Prompt tips
- Include business context and intended use
- List required fields explicitly
- Include expected data formats (dates, currency, IDs, status values)
- Mention whether relationships to other Tables are needed
Example:
Create a job applicant tracking table for a small recruiting team.
Required columns:
- Candidate Name (text)
- Email (text)
- Role Applied (text)
- Stage (single select: Applied, Screen, Interview, Offer, Rejected)
- Source (single select: Referral, LinkedIn, Website, Other)
- Applied Date (date)
- Recruiter Owner (text)
AI Insert (per-table row fill)
What it does
Inside a specific Table, AI helps fill a new Row from a short description.
Best use cases
- Quick data entry from unstructured notes
- Reducing repetitive manual form typing
- Creating first drafts that users then review
Smart AI Insert (table detection + insert)
What it does
AI analyzes text or image context, picks the best matching Table, then maps values into columns.
Best use cases
- User does not know where data belongs
- Intake workflows from screenshots/photos
- Mixed input streams where manual routing is slow
Ask AI about my data
What it does
On a Table page, users ask natural-language questions. AI answers based on recent Row data.
Examples:
- "How many open tickets do we have?"
- "Summarize the last 20 customer submissions."
- "Which items are low stock?"
Practical limits
- Responses depend on the quality and consistency of stored data
- Very broad questions can return less precise answers
- AI should assist decisions, not replace required operational checks
Accuracy + safety guidelines
- Keep field names clear and unambiguous.
- Review AI-generated values before creating Rows.
- Use select/status columns for critical state tracking.
- Audit sensitive workflows with human approval steps.
- Improve prompts over time based on observed misses.
Operational controls (recommended)
- Define per-company AI usage limits
- Track AI operation logs for debugging and monitoring
- Add feedback loops for "good suggestion / bad suggestion"
Troubleshooting quick list
- AI mapped to wrong field
- Improve field naming clarity and include examples in prompt.
- AI chose wrong Table
- Add stronger context in input and reduce overlap between similarly named Tables.
- Answer quality is vague
- Ask narrower questions with explicit filters/time ranges.
Related docs