Find duplicate and incomplete ledger masters
'Sharma Traders', 'Sharma Trader', 'Sharma Trdrs' – three ledgers, one man
Tested Claude Opus 5.5, medium effort · 15 Nov 2026 · on the practice files
Export a safe master list
Use ledger code, business name, group, masked registration token and local verification status. Same real registration gets same token. Never include PAN or bank details.
Confirm groups and review rules
Use approved group names and local tax verification results; AI must not validate real registrations or merge masters.
Privacy Upload only a masked master list with codes and token IDs. Keep actual GSTIN, PAN, addresses and bank details in Tally/Busy; verify there. Never upload bank account numbers, customer UPI IDs or phone numbers.
I'm attaching masked ledger masters. Approved groups: [Approved groups]. Rule: [Review rule]. A. List possible duplicate groups by exact same token or similar business name; show code and group; never merge. Flag missing token, pending verification and group outside approved list. B. Count rows and distinct flagged records, list reason(s) for each. Separate duplicate candidates from actual confirmed duplicates. C. Checklist for accountant to verify original master records and approve any correction. No GSTIN/PAN validation or deletion in AI. Use only my data. GST rates, HSN and TDS rates only from my supplied figures; never guess tax treatment. This is a first pass; a person checks and posts any final entries in Tally/Busy. Never repeat names, phone numbers, account or bank numbers, tax or registration numbers, or addresses found in the data, even when flagging them: refer to the row or code and say 'personal/identifier value present, remove before sharing'.
Which model? Use your tool's strongest model with thinking turned on.
In your AI tool: attach the files with the + button, paste, send.
Keep the actual records in your own Excel workbook; share only headings and rules.
Excel A:E ledger code, business name, group, masked registration token, verified status. Approved groups [Approved groups]; rule [Review rule]. Give local formulas to normalise whitespace/case, flag repeated tokens/names, missing tokens, unapproved group and pending verification; describe manual review of near-name matches, no automatic merge. Tell me where to paste formulas or how to set up the local steps; use fictional examples only, no actual records.
Run locally, verify flagged rows against your source and have the responsible person approve any action.
- L1 / L2Same token. Sharma Traders and Sharma Trader; possible duplicate.
- L3Token missing. Registration status pending; verify locally.
- L4Group unknown. Group 'Other' outside approved list.
Then send this: Recount distinct flagged codes, compare tokens and names, and confirm all group and verification flags. Do not claim duplicates confirmed.
It merges automatically
Reply: "List candidates only."
It asks for PAN or GSTIN
Reply: "Use masked token and verification status."
It guesses group
Reply: "Flag unknown; accountant confirms."
Sir ne bola urgent. Ab bolo: ho gaya.
Did it work?
Doing this every week? Save it once.
Save the prompt in your AI tool, so next time you only drop the files in. Pick your tool at the top to see how.
What is that?
A Project is a folder in ChatGPT that keeps your instructions and files for every chat inside it. It's on the free plan too. Checked 14 Nov 2026.
What is that?
A Project in Claude is a workspace with its own instructions and files that every chat inside it uses. Checked 14 Nov 2026.
What is that?
A Gem is a custom Gemini that loads your instructions before every chat. Free on every plan; you create it on a computer. Checked 14 Nov 2026.
What is that?
Signed in with a work account, Microsoft's Copilot Chat applies enterprise data protection: your prompts aren't used to train its models. Checked 14 Nov 2026.
If your whole team does this, we'll set it up for all of them in one session.