Bring in your Tally files, bank statements or receipt photos, explore every Debit & Credit by year → month → date, and export straight back to Tally. AI-assisted, and it all runs in your browser.
Three ways in — a Tally / ledger file, a bank statement, or a photo of a receipt. Everything is read on your device and lands in the same ledger below. Password-protected PDFs are supported; the password never leaves your browser.
QuickBooks-style. Totals roll up by year → month → date → transaction. Click any row to drill in.
Income by source and expenses by category come straight from your ledger. Add what you own and owe below to complete a real net-worth statement — saved only in this browser.
Every parsed and manual line. Pick a year, type or month — the summary and the list below update to match. Then export from Step 4.
Every input row (uploaded, scanned or manual) comes back out in Tally's own Day Book export format. Set the ledgers used to build bank vouchers, then download.
Date,Voucher Type,Voucher No.,Ledger Name,Debit,Credit,Narration
01-Jun-2025,Payment,1,Utilities,850.00,,DEWA electricity
,,,ADCB Current A/c,,850.00,
03-Jun-2025,Receipt,2,ADCB Current A/c,15000.00,,Client fee
,,,Professional Fees,,15000.00,
Date · Voucher Type · Voucher No. · Ledger Name · Debit · Credit · Narration), dates as DD-MMM-YYYY, VAT as its own Input VAT line. The Simple CSV is a one-line-per-voucher sheet that also feeds the Python main.py pipeline.Two ways to run it: manually (download the CSV/XML above and import in Tally), or directly from this page into TallyPrime running on your PC (Port 9000).
TallyPrime must be running with the gateway on Port 9000 (F1 → Settings → Connectivity → “Both”). For the direct button, run the tiny local bridge once.
TallyPrime → F1 → Settings → Connectivity (or F12 → Advanced).
Enable “Allow Server to use ODBC”. Accept & restart Tally.
Open http://localhost:9000 — you should see a Tally response.
Same pipeline as the buttons above, scriptable for batches. Extract → Claude classify → build XML → push. Files:
# 1 · install pip install -r requirements.txt # 2 · set your key export ANTHROPIC_API_KEY="sk-ant-…" # 3 · run the bridge so this page's buttons work python api_server.py # …or run a whole CSV batch straight into Tally python main.py entries.csv