Tax & Reconciliation

TDS & Form 26AS Reconciliation

Load your books and your Form 26AS. Get every counterparty resolved into matched, mismatched, books-only or 26AS-only — with the evidence behind each call, a chase list for the parties who under-reported, and working papers for the file.

A capability of ProcureTrail, configured by Kiren & Co as part of your professional engagement. It can be configured as part of an engagement on its own or alongside other ProcureTrail controls, depending on your organisation’s requirements. How the system fits together →

Sounds familiar?

📊

Your 26AS shows credits you cannot trace to an invoice, and the assessment is coming up.

📝

The same customer appears under four spellings and two TANs. Matching by name has already gone wrong once.

Reconciliation happens in a spreadsheet each quarter, and none of the working survives to the next one.

What the reconciliation does

A dedicated reconciliation engine, not a spreadsheet macro.

📂

Reads both 26AS formats

The portal spreadsheet and the caret-delimited text download are each parsed directly — headings, block structure, and transaction rows separated from summary rows.

🔑

TAN-anchored identity

Counterparties are resolved on the TAN. Spelling variants are held as aliases with a review queue, so the same customer under four names becomes one position.

Proposes, then waits for you

Automatic merging happens only on an exact TAN match. Name similarity proposes a link and a reviewer confirms it — because similarity alone was measured and found unsafe.

🎯

Four clear outcomes

Every counterparty lands in matched, mismatched, in-books-only or in-26AS-only, within a rupee tolerance you set.

🔍

A reviewer workbench

Work counterparty by counterparty, see the evidence behind each proposed match, confirm or revert it, and reopen anything that needs a second look.

📋

Outputs that finish the job

A reconciliation statement, a chase list, working papers and an outcome workbook — then finalise the run to fix the position.

See a sample reconciliation

We will walk you through a complete run on sample data — intake, alias review, the four buckets, and the workbooks that come out.

Reconciliation statement

The period position by counterparty, with the tolerance applied.

Chase list

The counterparties to follow up, ready to send.

Working papers & outcome workbook

The supporting detail behind every call, for the file.

Why the matching rules are conservative

Anchored on TAN

The TAN is the one reliable key across a client’s counterparties. Identity is built on it rather than on names.

Similarity proposes only

A similarity-based automatic merge was tested and produced two plausible but wrong merges against eight correct ones. It now proposes and a reviewer decides.

A bounded tolerance

The match tolerance is set in paise and is bounded, so an over-generous value cannot quietly declare every difference matched.

Every act recorded

Intake, alias confirmations, matches, reverts and finalisation are attributed and logged.

Often deployed with

Turned on per organisation. Same register, same audit trail, same login — no second system, no migration.

Common questions

What exactly is reconciled?

The TDS credit appearing in your Form 26AS against the corresponding entries in your books, counterparty by counterparty, for a period you choose. Every counterparty ends in one of four states: matched, mismatched, in your books but not in 26AS, or in 26AS but not in your books.

Which Form 26AS formats can be loaded?

Both. The portal spreadsheet download and the caret-delimited text download are each parsed directly, including the part headings, the block structure and the distinction between summary rows and transaction rows.

How are counterparties matched when the names do not agree?

Identity is anchored on the TAN, which is the reliable key. Where names differ across spellings and entities, aliases are proposed and a reviewer confirms them. Automatic merging happens only on an exact TAN match — name similarity alone will propose a link, never apply one.

Why does name similarity not merge automatically?

Because it was measured and found unsafe. In testing, a similarity-based exact-match layer produced two plausible but wrong merges against eight correct ones. A wrong merge silently misstates a counterparty’s position, so the software proposes and a person decides.

What comes out at the end?

A reconciliation statement, a chase list of counterparties to follow up, working papers, and an outcome workbook — then the run is finalised so the position is fixed.

Can we use this without the rest of Oparc?

Yes. TDS and 26AS reconciliation can be configured as the only thing your organisation sees — it has its own access controls and its own module switch.

Is this the TDS we deduct, or the TDS deducted from us?

The TDS deducted from you by your customers, reconciled against your Form 26AS. It is a separate matter from the TDS your organisation deducts when it pays vendors.

Want to see it run on your period?

We will show you a complete reconciliation on sample data first, so you can judge the method before any of your data moves.