AI chat assistant for your finances

You type a question in the app, and the assistant looks up the relevant transactions and receipts to answer it. In V1 it is read-only: it can summarize and explain your data, but it cannot add, edit, delete, or move anything.

The assistant is a third-party Azure AI Foundry agent, not a proprietary Quaestor model. It does not query your database directly — instead it calls a small set of read-only tools that are scoped to your own organization, so it can only see your data.

Each conversation is scoped by a signed session token, which keeps the agent confined to the authenticated account. The agent treats your ledger, receipt, and OCR text as untrusted data rather than instructions, so text that arrives inside an imported receipt cannot redirect the assistant.

The assistant answers concrete questions about the money already in your ledger. Useful starting points:

• Spending: “How much did I spend on groceries last month?” or “What were my five largest expenses in May?”
• Cash flow: “What was my net cash flow last quarter?” or “Did I spend more than I earned in April?”
• Subscriptions: “Which recurring charges do I have?” or “What am I paying every month for software?”
• Transfers: “Show the transfers between my checking and savings” or “Which moves between my accounts were excluded from spending?”
• Receipts: “What was on my last hardware-store receipt?” or “How much tax and tip did I pay at restaurants this month?”

Quaestor does not train its own models on your data. The chat assistant relies on a third-party AI provider, and that provider is configured to restrict training use where the option is available.

Provider data retention varies and is not something we can describe as zero-retention. For the current details on how chat and other data are handled, see our Privacy Policy.

Within the app, the assistant only ever reads your own organization’s data through read-only tools scoped per session — there is no path for it to read another customer’s data or to change yours.

People who would rather ask a question than build a report. If you keep months of transactions and receipts in Quaestor Ledger and want a fast, plain-language way to check spending, find a charge, or understand where cash went, the assistant turns that into a conversation — while staying read-only and scoped to your own data.