Skip to content

How it works

The work arrives prepared. You decide.

How routine work moves through Dhaaga on its own, the four rules that make it safe to leave running, and the exact count of what runs by itself today.

One ERP. Five connected ways of working.

CRM, production, purchase, sales and AI chat do not live in separate islands. These are the screens you work in, and the same business record moves through them from the first decision to the final entry.

Full flow in view
The Dhaaga CRM dashboard: open pipeline ₹4.3 L, weighted ₹1.2 L, five deals in play, a stage-by-stage funnel and a chart of where the pipeline came from — WhatsApp, phone, exhibition, walk-in and IndiaMART.
Lead to wonWorkflow live

CRM that keeps the next step visible

Keep enquiries, qualification, follow-ups, quotations and customer context in one connected pipeline instead of scattered chats and sheets.

The Dhaaga production dashboard: batches running, batches past due, value on the floor and output against target, with a daily output chart and where each batch is.
Plan to stock inWorkflow live

Production from BOM to finished stock

Connect material issue, work in progress, quality checks and finished stock so planning sees the same quantities as the floor and the books.

A Dhaaga purchase order PO-0007 to Gupta Hardware: partly received, with buttons to make the goods receipt, email the supplier or close the PO, and the supplier, dates, approval and total.
Request to payableIntake live · matching rolling out

Purchase with fewer manual checks

Move from indent and PO to receipt and supplier bill. AI document intake is live; automated three-way matching is shown as the next connected step.

A Dhaaga sales order for Sethi Stores: its steps from draft to dispatched, the item with HSN, tax and grand total, how much has been dispatched, and the customer’s balance.
Quote to cashCore flow live · follow-up rolling out

Sales connected all the way to cash

Carry the same record through quotation, order, stock reservation, dispatch, GST invoice and payment. Automated collections follow-up is rolling out.

Screens from the app, shown with sample data from a demo company.

An ERP records what you did. This one does the groundwork.

The difference is not a feature list. It is which side of the desk the work sits on — and therefore what you open in the morning.

200 forms becomes 1 inbox
The Needs your attention list on the Dhaaga Today screen: parties overdue, payments due today, orders still open, items to reorder, jobs running late and workers to be paid, each with an Open button.
The operating modelEverything routine moves. Only exceptions stop.That is how two hundred forms become one focused inbox.
Running a business on an ordinary ERPRunning it on Dhaaga
Someone types the supplier bill into the systemPhotograph the bill; it is read into the entry and you confirm it
Someone ticks the bank statement against invoicesLines are matched; only the ones that do not tie come to you
You find out about a stock-out when a sale failsThe reorder is proposed eleven days before the stock runs out
The overcharge is found at audit, months laterThe purchase above your contract rate is flagged the same hour
Month-end is a week of chasing missing paperworkThe books are current every evening; month-end is a review
You ask a person to pull a report and waitYou ask the system a question and get the computed number

Automatic is only useful if it is safe to leave alone.

Four rules the system is built to, not four things we promise to be careful about. An agent earns the right to act alone by its measured score; until then it drafts, and you approve.

4 rules
A Dhaaga purchase order with a Why? link under every line and its process lineage — indent, purchase order, goods receipt, quality check, supplier bill and settlement.
Trust by designEvidence in. Reversible action out.Every automatic decision keeps its source and its way back.
  1. The engine decides the number, the model decides the word

    Amounts, quantities, rates, dates, accounts and tax come from ordinary code with tests pinning it. A language model chooses wording and classification only. A model having a bad day gives you a badly worded note, never a wrong debit.

  2. Nothing is done without evidence

    Every action stores what it read to reach it — the document, the line, the rule that applied. Open any entry the system made and the reason is attached to it, not reconstructed afterwards.

  3. Every action can be undone

    No automatic action is allowed into the system without a declared reverse, tested as a round trip. Where an exact reverse is impossible the system books a compensating entry with a reason. Nothing is ever silently deleted.

  4. Switch the model off and your books still balance

    Without a model Dhaaga does less for you and asks more of you. It does not become wrong. A system that gives you incorrect numbers when the model is unavailable was never AI-native, only AI-dependent.

Try it

An agent posted this bill. Take it back.

Purchase bill INV-4471 · Sharma TradersPosted by agent
  • DrPurchases₹1,05,492
  • DrInput GST₹19,008
  • CrSharma Traders₹1,24,500

Here is exactly what runs by itself.

Every ERP in this market says its AI “handles your work”. We would rather hand you the count and let you check it. These numbers are measured from the shipped build, not from a roadmap.

Autonomy ledger
Measured on every releaseGated, not promised

An agent must clear an accuracy gate against 30+ real cases before it may act without a human. A score that falls fails the build.

Agents built, scored and pinned to a baseline

Every one runs its golden test set on every build, and a score that falls fails the build. Across them, 2,879 cases — 85 of the 89 score a perfect precision and recall in both judge modes.

89
Running fully unattendedLive

Cash count — tells you what should be in the drawer and what is not. It is deliberately read-only: it flags, and it can never post to your books.

1and read-only
Suggest the action, you approve itNeeds approval

Reorder watch, dead stock, rate contracts, three-way match, bank statement matching, payslip variance, books-vs-documents drift, advances, attrition risk and more.

88
Actions taken without stored evidence

If it cannot show the document, the line and the rule that applied, it is not permitted to act at all.

0
Wrong amounts caused by a language model

Amounts, quantities, rates, dates, accounts and tax come from ordinary code with tests pinning them. A model chooses wording and classification only — a bad model day gives you a badly worded note, never a wrong debit.

0
The fastest honest test

Put one real day through it.

Not a sandbox with sample data — your own bills, your own items, your own customers. By evening you will know whether this runs your business.

A full business day flowing through Dhaaga from opening tasks to balanced books