01
Choose your goal
Choose whether you want to investigate an issue, review SQL or PySpark, analyze historical data, or design a historical model.
← Back to Historical Data Assistant
HOW TO
How to use Historical Data Assistant
Start with your goal. The assistant guides the investigation, requests evidence only when it is useful, and turns the result into clear next steps.
02
Describe the problem
Explain what changed, what you want reviewed, or what you need to design. Start with your goal rather than a technical feature.
03
Answer focused questions
The assistant asks the smallest useful set of questions needed to understand the historical context and reduce uncertainty.
04
Provide evidence when useful
Add SQL, PySpark, CSV data, tables, or an image when requested. Evidence is used only when it helps verify a possible explanation.
05
Review the result
The result explains the likely root cause, confidence, supporting evidence, business impact, and recommended next steps.
PRIVACY
Customer evidence is not retained as product analytics
Raw SQL, PySpark, uploaded files, schemas, table rows and customer-specific implementation content are not stored for product learning. Only anonymized and structured metadata may be used to improve the experience.