State Bank of India statement PDF to Excel & Tally XML

Convert State Bank of India (SBI) bank statement PDFs with Jemsky’s dedicated parser — not a renamed homepage and not generic OCR. This page documents the layouts we actually ship, the quirks that break other converters, and examples CAs see in the wild.

State Bank of India e-statements are the most common PDFs Indian CA firms upload to Jemsky. They are not a single template: retail Yono downloads, branch-printed standard statements, and password-locked multi-month files all share SBI branding but differ in column geometry.

Jemsky’s SBI engine first classifies the layout from header tokens such as Customer Name, CIF No., and Ref/Cheque. Standard statements map debit/credit by coordinate bands; retail statements use midpoint classification so amounts that sit between 370–500 pt are not dumped into narration.

The practical CA problem on SBI is multi-line UPI and NEFT wrapping. A single transfer can occupy three visual lines. Generic OCR splits those into extra rows and breaks Opening + Deposits − Withdrawals = Closing. Jemsky groups preceding and succeeding narration onto the dated amount row, then verifies running-balance continuity across continuation pages.

Exports are Excel (.xlsx) with typed debit/credit columns, CSV, and TallyPrime Payment/Receipt XML. Nothing from the SBI PDF is stored after conversion. For party ledgers, import Master.json at Tally Party Ledger Masters and enable ledger suggest so Zomato/UPI rows do not all hit Suspense A/c.

Statement formats the SBI parser recognizes

• SBI Standard: Txn Date, Value Date, Description, Ref No./Cheque No., Debit, Credit, Balance
• SBI Retail: Customer Name / CIF No. / Available Balance headers with Ref/Cheque columns

Parser quirks specific to SBI

• Retail and standard layouts use different debit/credit midpoint bounds (retail debit mid < 435 pt).
• Narration fragments often sit above the date row (NEFT*ICIC…) or wrap below it; both are stitched into one particulars cell.
• Blank particulars on quarterly interest credits (typically 25 Mar/Jun/Sep/Dec) are labelled CREDIT INTEREST.
• Password-protected Yono / net-banking PDFs unlock in RAM only.

Example narrations and how they export

• UPI-MERCHANT/XXXX@ybl/123456 — UPI lines frequently wrap; parser joins them before Tally party matching.
• NEFT*ICIC0000123*YOUR CLIENT — NEFT refs are isolated from the wrapped description.
• CREDIT INTEREST — Quarterly interest rows with empty particulars keep a usable narration.

CA workflow on SBI files

For SBI month-end: convert the full multi-page PDF (not screenshot pages), confirm the audit equation matches the printed closing balance, export Excel for the working paper, then Tally XML with saved narration rules for recurring UPI merchants. Long retail files are where running-balance continuity matters most.

Exports and privacy

Every successful convert can download Excel (.xlsx), CSV, and TallyPrime / Tally.ERP 9 Payment and Receipt XML. Customer statement bytes never land in Jemsky’s database. Optional Tally Party Ledger Masters store ledger names you import — not this bank’s PDF.

1. Open the converter and upload the SBI PDF.
2. Leave Auto Detect on unless you know the exact layout name.
3. Confirm the financial audit equation matches the printed closing balance.
4. Export Excel for the working paper and/or Tally XML.

Frequently asked questions

How do I convert a State Bank of India statement PDF to Excel?

Upload the SBI e-statement on jemsky.com/app. Auto Detect should pick the SBI plugin. Review Opening + Deposits − Withdrawals = Closing, then download Excel, CSV, or Tally XML. The PDF is parsed in RAM and discarded.

Can I import State Bank of India statements into TallyPrime?

Yes. Export Tally XML and use Import → Transactions. Optional: import Master.json/xml and enable ledger suggest so SBI UPI/NEFT rows map to party ledgers instead of Suspense A/c.

Does Jemsky store State Bank of India statement data?

No. State Bank of India PDFs are not archived. Passwords unlock in memory only.

What is unique about the SBI parser versus other banks?

Retail and standard layouts use different debit/credit midpoint bounds (retail debit mid < 435 pt). For SBI month-end: convert the full multi-page PDF (not screenshot pages), confirm the audit equation matches the printed closing balance, export Excel for the working paper, then Tally XML with saved narration rules for recurring UPI merchants. Long retail files are where running-balance continuity matters most.

Related: All banks · User Guide · Security · Free Excel · Tally