Books and 26AS agree — same TAN, same section, same FY, within tolerance. Quarterly drill so you see where each credit was reported.
The credit exists on both sides but the amount differs beyond tolerance — a part-deduction, a rate gap, or a quarter booked to the wrong period. Quantified and traced to the voucher.
Credit on the 26AS that the return never claimed. Each line is quantified in rupees and traced to the deduction behind it, then flagged as potentially recoverable under s.199 — subject to the corresponding income being offered to tax (Rule 37BA) and the return not being time-barred for revision under s.139(5). Lines that belong to another year or assessee, or are a timing difference, are separated out. A reviewable position to confirm, not money in hand.
In your books but not yet on the 26AS — the deductor hasn't filed, filed under the wrong TAN, or there's a name mismatch on a TAN-less ERP. Name-bridging surfaces it instead of silently dropping it, and flags low-confidence matches for review rather than asserting them.
Books TDS receivable vs 26AS at (TAN, section, FY) grain with a quarterly drill and rupee tolerance. Name-bridging reconciles TAN-less ERP entries against the 26AS party — so a missing TAN doesn't break the match — and flags ambiguous or low-confidence name matches as their own review bucket.
Every 26AS credit absent from the return is quantified in rupees and traced to the voucher, then tested against s.199 / Rule 37BA — flagged as potentially recoverable only where the corresponding income is offered to tax, and held back where the line is time-barred, belongs to another year or assessee, or is a timing difference. It lands as an audit observation on the TDS-receivable balance and a 3CD input for management to act on — not a refund promise to hand the client.
FY-aware section-threshold breach testing, Rule 30 deposit due dates, s.201(1A) interest on late deposit and s.234E late-filing fee — each flag carrying its section and the period it applies to, and a check is suppressed when the provision is not in force for the year under audit.
Books deductions vs the 24Q and 26Q filings, surfacing deductions made but not reported, and reported amounts not deposited — so a filing gap is caught in the audit, not in a notice.
Books TCS collected vs 27EQ, sorted into matched, unfiled, unbooked and not-collected — with s.271CA exposure surfaced on the not-collected lines. Form 3CD clause 34(a)/(b)/(c) is auto-filled from the full TDS and TCS tie-out across 24Q, 26Q and 27EQ, not from any single form.
TCS borne as a buyer read from 26AS Part V, with a 206C(1H) threshold engine that tracks the ₹50 lakh per-buyer trigger across the FY and flags collection that should have applied.
Reads the TRACES 26AS (text or PDF) and the 27EQ / 24Q / 26Q consolidated files alongside your Tally, SAP or Excel TDS-ledger export. Blank or garbled PANs and TANs, deductors filing under the wrong TAN, amended TDS statements, part-quarter deductions and TAN-less ERP entries are name-bridged and tolerance-matched, with the unresolved rows routed to a review bucket rather than silently dropped. Built for an actual 26AS, not a clean CSV.
Variances are tested against a materiality threshold under SA 320, and period-lag timing differences are auto-flagged — so your team reviews the credits that matter, not a rounding paisa or next quarter's deposit. You set the name-bridge confidence threshold and review borderline matches before they count.
Every match, variance and s.199 flag is timestamped, rule-cited and traced to its source row, with the 26AS snapshot pinned — exportable to your working-paper file and re-performable under SA 230. Deterministic: the same inputs give the same result every run.
Lock section mapping, tolerance and threshold treatment once at the firm level; every branch and every article runs identical logic, with overrides flagged for partner review. Inconsistent treatment between offices is what a peer review or NFRA file catches — this removes it.
No more VLOOKUP marathons against a 26AS PDF and a 27EQ dump. The match is done at TAN-section-FY grain; only the exceptions need a human.
Open the buckets, not the whole ledger. Every variance and every potential s.199 credit is quantified and traced to the deduction behind it, with low-confidence name matches called out so you can size the review before you commit the team.
The TDS credit claimed is tested for completeness and accuracy, the s.199 position is quantified as a 3CD input and an observation for management, clause 34 is filled from the working — and the whole chain is shown for peer review. The claim and any revised return stay the firm's call.
