Common questions
How accurate is the extraction?
We do not publish a headline accuracy rate. Every field carries its own confidence score, so anything below your threshold lands in the review queue instead of posting silently. Run the live demo β it shows the exact scores and checks for each document type, including a deliberately imperfect cafΓ© receipt.
Does the demo read photos, scans, or PDFs?
No. The public demo parses plain text only β use the text or OCR output of your document for the demo. Signed-in accounts can upload receipt images for OCR as a test; production image capture is set up during trial onboarding. See the demo guide.
Which languages and currencies does the demo handle?
English month names, ISO dates (2026-10-24), and day-first numeric dates (24/10/2026), plus three-letter currency codes on the total line. Other formats are not silently guessed β missing or unrecognised fields are routed to review. Full parser rules: documentation.
Where is our financial data stored?
Demo text is processed in memory for the result only and is not stored. Trial requests store only the contact details you submit. Data is encrypted in transit with TLS. See Privacy and Security.
What happens to the documents themselves?
Source files are kept with your account until you delete them, and can be purged on request. Structured records persist in your accounting system independently of our storage.
Do you support partial or bulk migration?
Historical archives can be backfilled from a shared drive or S3 bucket, and the pipeline applies the same validation rules so pre-existing duplicates surface during migration. This is an onboarding feature, set up per customer.
What does the trial include?
A trial workspace, help connecting your first source, and help connecting your first ledger. No card details are requested with a trial request. Start here: free trial.
Are the integrations live today?
Not from the public demo. Signed-in accounts can test a QuickBooks Sandbox push; the integrations page states each system's status honestly. Production connections are set up during trial onboarding with your own OAuth approval.
Is tvmmm for bookkeepers with many clients, or for one company?
Both, but it is positioned for bookkeepers first: one intake and one supplier-rule set per client, publishing to each client's own ledger. A single company uses the same pipeline with one client workspace. See the bookkeepers page.
What is the difference between receipt capture and expense management?
Receipt capture reads documents and publishes them into an accounting system. Expense management adds the steps around employee spending: submitting expenses, getting approval, reimbursing, and enforcing spending policies. Firms handling client bookkeeping usually need capture; companies whose own employees spend usually need both. tvmmm is a capture-first product.
How should a bookkeeper handle a client's receipt backlog?
Bulk-upload the backlog, let supplier rules code what they can, and review the rest in batches by supplier or month. Match documents to existing bank-feed entries instead of creating new ones, so nothing is recorded twice. Once the backlog is clear, switch the client to a single ongoing intake route.
Anything else: [email protected].