Live today NDTL computed from a real bank’s trial balance

Reconcile once.
File everything.

Every fortnight, four people spend hours rebuilding one return. By hand. Your core banking system already produces the data — Techarthik turns it into every regulatory and financial output your bank owes.

Reconcile once · Validate once · Reuse across the bank
techarthik / returns / statutory / 27-Mar
CRR & SLR compliance — computed
Derived from the imported trial balance · fortnightly cycle
Validated
NDTL — computed
₹44,40,96,78,114
Derived from a real trial balance. Every bucket exposed.
CRR required @ 3%₹1,33,22,90,343
CRR held with RBI₹1,40,00,00,000
CRR excess₹6,77,09,657
SLR required @ 18%₹7,99,37,42,060
SLR assets held₹11,80,28,20,773
SLR excess₹3,80,90,78,713
0
Workflows the bank already runs by hand — on one controlled dataset
0
RBI returns modelled end to end, each traced to its legal reference
0
Shared calculation engines — NDTL, IRACP/SMA, CRAR, ALM and more
0
Faster on a fortnightly return — 40 person-days become half a day
The product suite

Six workflows the bank runs by hand
— on one controlled dataset.

These are not six products. They are six workflows reading the same reconciled data — so a figure proved once is read by everything downstream. Select any module for the before-and-after.

Status is labelled honestly. Financial Statements is a compiled application in active use at a real co-operative bank. The RBI returns suite is validated against a real bank's trial balance and is not yet connected to RBI's CIMS portal.

Why it is needed

Today one figure is rebuilt by hand
for every return that needs it.

Today Manual

Core banking Trial balance Excel Manual mapping Reconcile Recheck Return

Five human hand-offs, repeated for each of 29 returns. NDTL is re-derived separately in every one — and nothing checks that they agree.

With Techarthik Controlled

Core banking Trial balance One controlled dataset Validate Return

Proved once, then read by everything downstream. NDTL is computed once; every return reads the same figure.

The data is not the problem. What happens to it afterwards is: no shared source of truth, and no cross-check between returns required to agree.

What it saves · one return, one cycle

Ten days of four people,
or half a day of one.

0
faster
40 person-days become half a working day
~₹90,000
saved per cycle
against ~₹90,000 per cycle in manual staff cost
~₹21 Lakh
saved / year
on Form B alone — one of twenty-nine returns

Basis. Today: 4 people × 10 days = 40 person-days, ~₹90,000 per cycle in staff cost. With Techarthik: one person, half a working day, ~₹1,100 per cycle. Calculated on an average salary of ₹50,000 per person per month. Indicative — Techarthik's own projection, not measured at a live client.

What the status quo costs

Co-operative banks are the most-penalised category the RBI regulates.

264
enforcement actions in FY 2024-25
More than any other category of regulated entity.
₹15.63 Cr
in monetary penalties
Exceeds private and public sector banks combined.
₹122 Cr
New India Co-operative Bank, Feb 2025
Alleged embezzlement; RBI froze withdrawals.

No software prevents fraud, and none of this claims to. The narrower point: weak processes have consequences, and this sector carries more of them than anyone else.
Source: RBI Annual Report FY25; RBI press release, 24 February 2025.

Why now

The rulebook changed.
All of it, in one day.

28 November 2025

One consolidated Master Direction

Capital · Governance · Credit · ALM · IRACP · CRR · SLR — restated together.

15 December 2025

The Form B fortnight itself was redefined

The reporting fortnight — the unit the whole return is built on — changed.

In a controlled system a rule change is a configuration update. In a workbook it is a rebuild.

The catalogue

29 returns. Six families.
All due-dated.

Each one carries its own statutory citation, applicability rule and frequency. Nothing invented — the field structures come from the RBI specifications and Master Directions themselves. Select any return for detail.

The filing cycle

Seven steps, one screen each

Every one of the twenty-nine returns runs the same workflow — the same import, the same engines, the same validation and the same trail. The team learns it once.

techarthik / returns
29returns, one dataset
7shared engines
0figures re-keyed
Under the hood

Regulatory maths, implemented as code

The hard part of a return is not the form — it is the computation behind it. Seven shared engines implement the RBI algorithms directly, so figures derive themselves from the ledger instead of being typed in from a side workbook.

Cross-validation

If two returns must agree,
the system checks.

Block

Form B NDTL must equal Form I NDTL

PASS
Warning

Form B bank credit must equal OSS-1 loans and advances

PASS
Block

SLR held must not fall below the statutory minimum

PASS

Block stops filing. Warning flags it. Either way it is caught inside the bank. Automation speeds a wrong number — control questions it.

Audit trail

Every change — who, when, what.

CreatedMaker · 10 Aug 2026
Imported trial balanceMaker · 10 Aug 2026
Imported inter-bank depositsMaker · 10 Aug 2026
ValidatedMaker · 10 Aug 2026
FinalizedMaker · 10 Aug 2026
FiledMaker · 10 Aug 2026
Validated againMaker · 12 Aug 2026

Four questions the spreadsheet cannot answer.

Who entered it?
When?
What changed?
On what basis?

Even a re-validation two days later is on the record. Twenty-eight more returns — same dataset, same trail.

The platform

Six workflows. One data layer.

RBI ReturnsCredit BureauNPA Financial StatementsFixed AssetsGST
One controlled data layer

Reconciled · Validated · Versioned · Traceable

Read by every function above it. A figure proved once is read by everything downstream, and returns that must agree are cross-checked before filing.

Works with what you already have

No rip-and-replace.

Techarthik reads from your existing core banking system. Your bank does not need to replace anything for it to work.

FinacleKiya.aiTCS BaNCSC-Edge Oracle FLEXCUBEIntellectWebsoftexSmartBooks
For the chairman, CEO and board

Five questions. Instant answers.

A low-risk way in

Start small. Prove it. Expand.

01

START

One workflow.

02

RUN

In parallel with your current process.

03

PROVE

The control, on your own numbers.

04

EXPAND

Across the bank.

No bank-wide migration.

See it move

The product, running

A short film on the platform — the six workflows, one controlled dataset, and what the bank stops doing by hand.

What we are building next

The road ahead

Shipped is shipped and in-build is in-build — we label the difference honestly, because our customers are regulated entities and so are their auditors.

In build

Live CIMS submission

Direct filing into RBI's CIMS portal with the official XBRL taxonomy, replacing today's taxonomy-ready export.

In build

Native CBS connectors

Scheduled pulls from the core banking system, so the trial balance and CASA extracts arrive without a manual export step.

In build

Fixed asset register

QR/barcode tagging, branch-wise register and automated depreciation with a full movement trail.

Next

GST reconciliation

GSTR-2A/2B versus purchase register matching, RCM tracking and s.17(4) ITC treatment.

Next

Board & MIS pack

Auto-generated board reporting pack drawn from the same reconciled dataset as the returns.

Exploring

Anomaly detection

Period-over-period variance flags on ledger movements and asset-quality slippage, surfaced before filing.

Let's automate your next filing cycle.

We take the manual work — and the risk — out of your bank's RBI reporting. A walkthrough runs about forty minutes, on your own return set.