Chat with your data is a private, read-only analyst for the active Intelligence scenario. It answers in Italian or English from existing transcripts, extractions, and metrics. On Growth and Scale it can also read completed, human-approved Evidence Synthesis findings.
The analyst answers questions about a frozen scenario scope. It cannot create or edit a scenario, campaign, respondent, Knowledge Base, or Evidence Synthesis run. Changing filters starts a new analysis; earlier sessions keep their original scope.
Substantive claims are labelled Exact data, Observation, or Interpretation and cite a server-issued source locator. Broad questions use a bounded sample and may not claim population prevalence. No-evidence and failed answers add no credits.
Free sees the surface and a Pro upgrade path. Pro, Growth, and Scale receive the same active-scenario data scope. Evidence Synthesis findings remain Growth and Scale only; Pro is pointed to that upgrade, not given silent access through the analyst.
Ask grounded questions first. When you need emerging categories, contrary cases, and a reviewed report over a locked 10–20 transcript sample, open Evidence Synthesis. The analyst never starts that work and never replaces researcher judgment.
This beta does not invent a codebook, form theory, or declare statistical significance or theoretical saturation. Transcript text is treated as untrusted content. Successful grounded answers are shadow-metered at 5 credits during launch; the balance is not deducted.