How Vanteloq AI Uses Business Data and Privacy Controls

Separate Calculation From Interpretation

Business analysis has two jobs: calculate a result correctly and interpret what it means. Vanteloq keeps these roles distinct. Its report logic calculates measures from approved records. The AI layer receives a bounded summary and can explain the result, identify questions worth investigating and help a user navigate the application.

This distinction is useful when reviewing profit, sales changes or inventory. An explanation should be checked against the formula, dates and coverage shown in the report. A confident answer does not repair a missing cost or establish why customer behaviour changed. Treat the answer as analytical assistance that can be questioned and verified.

How Role-Based Access Controls Shared Business Data

When Workspace data is on, Vanteloq attaches permitted summaries relevant to business analysis. These can include financial KPIs, source status and authorized inventory, marketing or BookLoQ context. Access remains scoped to the workspace, the user's role and permitted locations. Enabling AI does not grant a teammate additional financial or customer-data permissions.

Automatically supplied evidence excludes credentials, full account numbers, customer identities, original invoice files and raw transaction lists. Your written question is still part of the request, so avoid pasting secrets or unnecessary personal information into it. A summary is a deliberate selection of evidence, not unrestricted access to the business database.

Manage AI Privacy and Conversation Memory

Workspace data controls whether business summaries accompany a question. With it off, a user can still ask for product guidance or an explanation of a financial concept without attaching workspace records. Conversation memory controls whether new chat messages are saved in Vanteloq and whether matching recent history can inform later replies.

Memory starts off. When enabled, up to six recent messages can be reused while the evidence and access permissions still match. Turning memory off stops future saving and history use, but it does not erase earlier saved chats. Use the saved-chat deletion controls for that purpose. Provider safety logs and managed backups have separate retention rules.

Ask a Question That Can Be Checked

Start from the report you are reviewing and select its location and dates. A focused question might be: 'Which recorded changes contributed to lower sales in this period, and what information would you check before changing prices?' This invites an explanation with boundaries, rather than an unsupported instruction to take action.

For BookLoQ, review the available ledger or cash information first. Then ask which recorded commitments affect a cash scenario and which assumptions need confirmation. Organization-wide BookLoQ analysis requires the appropriate finance access. If evidence is missing, ask what needs to be supplied instead of requesting a more confident version of the same answer.

• State the decision you are considering and the reporting scope.
• Ask the answer to distinguish recorded facts from assumptions.
• Check important figures and proposed actions against the source report.

Keep Approval With the People Responsible

Vanteloq AI does not post journals, pay bills, file tax returns, place purchases or contact customers. People review the explanation and decide what happens next. This matters because incomplete records, unusual transactions and local business context can materially change an otherwise plausible interpretation.

A suggestion can lead to a review task with a named owner and due date. Record the proposed action and the evidence that would support it. When reviewing the outcome, distinguish completing the task from demonstrating that it improved the business. Financial statements and consequential accounting decisions still need the appropriate professional review.

Know What the Public Demo Demonstrates

The public Vanteloq demo runs with fictional records and working calculation logic. It lets visitors inspect sales, source records, inventory questions and BookLoQ cash scenarios without signing in. Its evidence briefs are rule-based examples, and using the demo does not send a question to an AI provider.

The authenticated workspace is different: Vanteloq AI uses OpenAI after the applicable data-use notice and access checks. Try the demo to understand the reporting workflow, then review Help and AI Settings before using your own business information. Check the current privacy policy for processing locations, retention and deletion details.