LedgerBridge in 90 seconds.
How each accounting API connects, what you get out, and how much manual work disappears. Then watch it run on real TallyPrime.
Figures are from the demo company "XYZ Limited". Time savings are illustrative estimates and depend on volume and data quality.
Real Tally and QuickBooks. Real output. No mock-ups.
Recorded on a Windows PC running TallyPrime. The video shows the company in Tally, the actual API request and response, the commands that ran, and the Excel files LedgerBridge produced from the live data.
And on QuickBooks Online, through the official API
The same PC, connected with OAuth to an Intuit QuickBooks Online sandbox company with fictional US books. You see QuickBooks' own trial balance, the API calls, the commands, and a lender pack in US dollars.
New: the Cash & MSME cockpit, on the same live books
Recorded 10 Nov 2026. A practical Monday-morning scenario: the CFO asks where cash is stuck. We walk Tally's debtors and creditors, run one command against the live company, then open the dashboard and Excel workbook it produced. Then the same on QuickBooks Online, in US dollars.
Open the Tally dashboard QuickBooks version
Recorded in TallyPrime Educational mode, which accepts vouchers only on the 1st, 2nd or 31st of a month, so voucher dates sit on the 1st and 2nd. The figures are unaffected. Want the full walkthrough? Watch the complete 7-minute tutorial.
What actually happens on a Tally PC
The same four steps you see in the video. Every screenshot below is a frame from the recording.
Turn on Tally's API
In TallyPrime: F1 Help › Settings › Connectivity › Client/Server configuration. Set it to Both on port 9000. Nothing is installed inside Tally.
Pull the books
One request returns every ledger with group, balance and party GSTIN. LedgerBridge walks Tally's group tree, so loans, debtors and creditors are classified correctly.
python -m ledgerbridge.agent export --company "XYZ Limited" --from 2026-04-01 --to 2026-09-30 --out books.json
Reconcile GSTR-2B
Purchase vouchers are matched to the 2B JSON from the GST portal: matched, fuzzy match, value mismatch, missing, ineligible. You get an IMS action for each invoice and a supplier chase list.
python -m ledgerbridge.agent recon --company "XYZ Limited" --gstr2b 2B.json --out recon.xlsx
Build the lender pack
P&L and balance sheet in ₹ crore, DSCR and leverage, receivable and payable ageing, GST turnover against books, and the trial balance in absolute rupees for the audit trail.
python -m ledgerbridge.agent pack --company "XYZ Limited" --from 2026-04-01 --to 2026-09-30 --out pack.xlsxQuickBooks Online (shown in the second video), Xero, Zoho Books, Business Central and NetSuite follow the same four steps. Only step 1 changes: you approve access once with OAuth instead of opening a port.
How each API connects
Connect once per company. After that, books sync nightly without anyone exporting a file.
| Accounting system | What you provide | How LedgerBridge connects | Typical setup |
|---|---|---|---|
| TallyPrime / Tally.ERP 9 | Tally open on the office PC with port 9000 enabled | Desktop agent reads ledgers and vouchers over Tally's XML interface | 10 min |
| QuickBooks Online | Approve access once (OAuth) | Intuit Accounting API: accounts, trial balance, bills, invoices | 5 min |
| QuickBooks Desktop | Windows PC with QuickBooks; authorise LedgerBridge once | qbXML through the QuickBooks SDK | 15 min |
| Xero | Approve access once (OAuth) | Accounting API 2.0: accounts, trial balance, invoices, contacts | 5 min |
| Zoho Books | Approve access once (OAuth) | REST v3: chart of accounts, trial balance, bills, invoices | 5 min |
| Odoo 16–18 | Instance URL, database name and API key | JSON-RPC: accounts, journal items, vendor bills, customer invoices | 5 min |
| Dynamics 365 Business Central | An Entra ID app registration | API v2.0: accounts, trial balance, purchase and sales invoices | 20 min |
| Oracle NetSuite | Token-based auth with a read-only role | SuiteQL over REST | 20 min |
| SAP, Sage, MYOB, Busy, Marg… | Trial balance and invoice list (CSV or Excel) | Universal importer with column matching and account classification | 10 min |
Setup times are typical estimates for a first connection.
Exactly what comes out
Every output is computed by rules and ties back to vouchers. AI only drafts the emails and commentary.
Clean ledger data
JSON over REST: ledgers with groups and balances, vouchers with every leg, party tax IDs, and a trial balance that always balances. The same schema for every source.
- Replaces manual exports to Excel
- Feeds your own models, BI or lending system
GST reconciliation workpaper
Excel with three tabs (Summary, Invoice detail, Supplier follow-up): a status and an IMS action per invoice, ITC at risk versus safe to claim, and drafted supplier emails.
- Replaces line-by-line 2B matching
- Catches format-only mismatches such as MP-421 vs MP/0421
Lender Pack
Excel with six tabs: summary in ₹ crore, ratios with DSCR, receivable and payable ageing, GST turnover against books, and the trial balance in absolute rupees, plus flags for the credit desk.
- Replaces rebuilding ratios by hand
- Weak points are flagged, not hidden
Per company, per month
Your team reviews exceptions instead of typing data.
| Task | Manual | With LedgerBridge | What is automated |
|---|---|---|---|
| Extract ledgers | 2 h | 2 min | Nightly sync from the source system |
| GSTR-2B reconciliation | 6 h | 5 min | Matching, fuzzy invoice numbers, IMS actions |
| Supplier follow-ups | 2 h | 10 min | Drafted emails grouped by supplier |
| Lender pack and ratios | 8 h | 15 min | P&L, balance sheet, DSCR, ageing, flags |
| Total | 18 h | 32 min | About 97% less manual work |
Illustrative estimates for a mid-size company file; your figures depend on invoice volume and data quality.
Set it once, then it runs every month
- Schedule the nightly sync with Windows Task Scheduler (Tally, QuickBooks Desktop) or let the cloud pull run for OAuth sources
- Call the API from your own systems to pull reconciliations or packs on the 11th of every month
- Ask an AI assistant through the MCP connector, for example "What is our ITC at risk this month?" or "Build a lender pack as of 30 September"
# nightly: sync books from Tally python -m ledgerbridge.agent sync --company "XYZ Limited" # monthly: reconcile GSTR-2B to Excel curl -X POST $LEDGERBRIDGE_API/v1/companies/42/gst/reconcile?format=xlsx \ -H "X-API-Key: $KEY" -F gstr2b=@2B.json -o recon.xlsx # on demand: lender pack curl -X POST "$LEDGERBRIDGE_API/v1/companies/42/lender-pack?as_of=2026-09-30&format=xlsx" \ -H "X-API-Key: $KEY" -o pack.xlsx
Try it on your own books.
We connect your first five companies with you and run the first reconciliation together.