Straight answers.

Can we just sign up and try it, or do we need an invitation?

Yes. Open app.datavrn.com/login and enter your work email. The one-time code creates your account; no invitation or separate sign-up process is required.

Start with one standalone Entity: bring a Trial Balance, confirm the prescribed Schedule III groupings and disclosures, and export an Excel working draft with live links or a clean finalised workbook with fixed values. Design partnership is optional and separate from normal product access.

What is available today?

Today, one standalone Entity can take a Trial Balance through confirmed prescribed groupings and disclosures to an Excel working draft with live links or a clean finalised financial-statements workbook with fixed values in the prescribed Schedule III format. Datavrn prepares the statements; it does not file or certify them.

Broader governed management-reporting workflows—including allocation, consolidation, operating metrics and the monthly pack—are in development.

Which systems does Datavrn read from?

Three ways in, today. Spreadsheet uploads — canonical templates covering the trial balance, transactions, receivables and payables, masters, and the operating data around the ledger. Tally file exports — trial balance and day book, in the XML or Excel formats Tally already produces. And a Zoho Books connector that syncs over OAuth.

If your books live somewhere else, tell us — it informs what we build next.

What does setup actually look like?

You create your organisation, add your entities, and bring data in — the product suggests column mappings and account classifications, you confirm them, and your confirmations are saved so the next cycle is a re-upload, not a redo. Nothing is ever auto-applied.

Your confirmations carry forward, so later cycles begin from the governed method already established. Design partnership is optional and separate from normal product access.

What does a design partner get — and what do you expect?

Not access to the product — anyone can sign up for that. What a design partner gets is a direct line to the two founders and a real say in what gets built: the methodology you have refined over years, encoded and running as software. We expect you to run it on genuine reporting cycles, tell us directly where the workflow falls short, and help validate it against the complexity of real finance operations. Commercial terms are agreed individually with each partner.

Where does our data live?

Your financial and operational reporting data is stored in India. The database runs in Mumbai (AWS ap-south-1), and your organisation’s data is isolated from every other organisation’s. Datavrn sends none of that reporting data to an AI provider unless your organisation explicitly enables and uses a feature that requires it. More detail is on our security page.

Can we get our data back out?

Yes — the design principle is that nothing is held hostage. Datavrn preserves source provenance and the source fields permitted for each format; people-data formats intentionally exclude personal fields. Supported ingestions can be rolled back, but rollback removes that upload and does not always restore rows replaced by an earlier upload. Available mappings export and round-trip as standard files; report output includes linked Excel. Deleted Entities are hidden and recoverable, and the action is audited.

Can we use Datavrn without AI—or work through our own AI?

Yes. Every reporting workflow currently available in Datavrn can be operated without AI. If your team prefers to work conversationally, connect your enterprise AI to supported Datavrn tools or talk to optional Datavrn AI. No AI provider is required to use any released Datavrn workflow.

Through your enterprise AI, the team can ask for information and carry out supported Datavrn actions within the permissions you grant it. With Datavrn AI, users can talk to Datavrn to find their way and understand supported reporting results. The conversation changes how you interact with Datavrn; it never changes how a reported figure is computed.

Every reported figure is produced by a deterministic engine. No model generates, estimates or adjusts one.

Your enterprise AI works through scoped, permission-bound access and records actions on behalf of a named human principal. Data it requests passes through your chosen AI provider under your agreement with them. For Datavrn-operated Help, your submitted question passes to a disclosed AI provider; Datavrn does not attach reporting data. Prepare sends the question you submit plus bounded, allow-listed reporting context and facts from one selected frozen statement version, only for a user-initiated request after organisation enablement and explicit Entity consent.

Does Datavrn replace our accountant or analyst?

No — it removes their plumbing. The judgement stays with your team: how costs should be apportioned, what a variance means, what the board should be told. Datavrn encodes those decisions once and executes them every cycle, with the working shown, so the people who understand the business spend their time on the decision rather than the assembly.

Who is behind Datavrn?

Two Chartered Accountants, both qualified in 2013, who spent their careers producing exactly this kind of reporting before deciding to build it as software. The longer version is on the about page.