UNDERWRITING, WITH THE EVIDENCE ATTACHED

Underwrite the project.
Not the paperwork.

From messy developer Excel models and loan documents to source-linked financial inputs, reliable project cash flows and credit-ready risk analysis. Built around how Indian real-estate lenders actually work.

Human-in-control underwriting. No black-box loan approvals.
RiskSarthi / Deal workspace
concept
RESIDENTIAL CONSTRUCTION FINANCE · PUNE

Green Heights Residency

REVIEW OPEN
Mapped model inputs 4 source records
Loan facilitySanction letter · page 03
₹70 Cr
Project costProject model · G42
₹150 Cr
Booked salesSales register · H85
₹80 Cr
CollectionsBank reconciliation · pending
₹32 Cr
Facility loan-to-cost
46.7%Traceable formula
Cost to complete
₹85 CrIllustrative forecast
Monthly cash-flow projectionILLUSTRATIVE
Input snapshot · Concept caseIllustrative product interface — not a real credit assessment
Every critical figure.
Back to the source.
THE UNDERWRITING BOTTLENECK

Important credit decisions are still buried in scattered files.

When a developer's sales register, construction budget, valuation report and sanction letter disagree, analysts spend their time finding and reconciling numbers before they can assess the risk.

Every developer model is different.

Columns change, units switch between lakh and crore, formulas break, and the same number appears across multiple workbooks.

01 / DATA FRAGMENTATION

“Where did this figure come from?”

A credit reviewer needs the original cell, clause or certified document—not just an AI-generated number with a confidence score.

02 / EVIDENCE GAP

Profitability can hide liquidity risk.

Booked sales are not collections. An apparently profitable project may still face a cash shortfall before completion.

03 / CREDIT EXPOSURE
RiskSarthi connects the documents, the numbers and the decision.

Less spreadsheet reconstruction. More time assessing cost-to-complete, repayment capacity and exceptions.

See the workflow →
THE DIFFERENCE

Click the number.
See the evidence.

RiskSarthi is being designed around a verified mapping grid. Every source observation carries a file, location, raw value and review status. Explore an illustrative example below.

Financial mapping grid

Green Heights · 4 example variables
CONCEPT
Financial fieldValueSourceStatus
Sanctioned loan₹70 CrPDF · P03Verified

Select any row to inspect its original location.

Source inspector

PDF / PAGE 03
Sanction_Letter.pdf›Page 03
ORIGINAL EXTRACTED EVIDENCE
Sanctioned Facility: ₹70,00,00,000
Mapped variableloan.facility_amount
Normalized₹70 Cr
Source observation → mapped variable → analyst verificationDemo data only. No real documents are uploaded here.
ONE CONNECTED UNDERWRITING WORKBENCH

From raw files to a defensible credit review.

Six connected capabilities, designed specifically for development-finance workflows rather than generic property dashboards.

Document-to-model ingestion

Map Excel, PDFs and supporting project records into standardized financial fields. Preserve original values and evidence.

Extract • Map • Review

Two-way source traceability

Navigate from a calculation to its input, original workbook cell or PDF clause—and back again.

Provenance • Versions

Development cash-flow engine

Model project spending, receivables, collections, promoter equity, loan drawdowns, interest and funding gaps.

Monthly • Deterministic

Stress testing that explains

Test sales delays, price reductions, construction overruns and interest-rate shifts, then follow the liquidity impact.

Downside • Sensitivity

Lender policy and exceptions

Surface lender-configured LTC, equity, disbursement and evidence exceptions without issuing autonomous loan decisions.

Policy • Human review

Source-linked credit memo

Prepare structured underwriting summaries with calculated figures, scenario drivers, outstanding questions and review history.

Credit appraisal • Audit trail
INTERACTIVE PRODUCT CONCEPT

Know where the cash breaks. Before the project does.

Project profit alone cannot answer whether construction can be funded through completion. Model the timing of collections, spending, promoter funding and eligible debt drawdowns.

  • Identify liquidity pressure early
  • Compare base and downside assumptions
  • Explain which inputs drive the deficit

Explore a liquidity downside

Three-month hypothetical cash budget · ₹ crore
₹8 Cr
₹3 Cr
₹0 Cr
Projected closing cash−₹9 Cr
Period-end funding shortfall₹9 Cr

This is a simplified, illustrative calculator, not a verified underwriting output. Full calculations must use period-by-period eligible cash, contractual draw conditions, interest and repayment schedules.

HOW IS DESIGNED TO WORK

A clear chain from information to judgement.

01 / SOURCE

Upload the deal pack

Developer workbook, sanction letter, cost reports, sales and collection MIS, technical and valuation evidence.

02 / VERIFY

Reconcile every important input

AI suggests; the analyst accepts, edits or flags conflicts. Evidence stays attached to each accepted value.

03 / ANALYSE

Run the financial model

Calculate monthly cash flows, cost-to-complete, debt service, repayment timing, scenarios and policy exceptions.

04 / REVIEW

Explain and document the risk

Create source-linked credit outputs, preserve assumptions and record the authorized review history.

It is an underwriting decision-support layer—not an autonomous credit approver.
Integrate with existing Excel and lender workflows before replacing them.
Discuss your underwriting workflow
BUILT FOR INDIA'S CREDIT WORKFLOWS

Purpose-built for the people evaluating development risk.

starts with Indian residential development finance: cost-to-complete, promoter funding, construction milestones, pre-sales, escrow-aware collections and structured credit appraisal.

01 / INSTITUTIONS

Real-estate NBFC teams

Evaluate residential construction finance proposals and scrutinize project economics, collateral, equity and repayment capacity.

Construction finance • Credit appraisal • Loan monitoring
02 / CAPITAL

Private credit & debt funds

Compare financial models and lender cases with consistent assumptions, scenario testing and traceable investment committee materials.

Real-estate debt • Structured credit • Downside analysis
03 / PROFESSIONALS

Credit analysts & reviewers

Spend less effort reconstructing numbers and more time investigating anomalies, underwriting risks and exception rationale.

Excel spreading • Model audit • Credit memos
QUESTIONS WE EXPECT

Built for scrutiny, not just a slick demo.

We are treating as a verification-first product. These are deliberate product boundaries, not hidden limitations.

Does RiskSarthi approve or reject real-estate loans?

No. The proposed platform is decision support. It highlights verified financial outputs, policy exceptions and downside risks; authorized lender personnel retain credit decision authority.

Will it replace our Excel models?

The initial goal is to ingest existing Excel and PDF files, link the mapped figures to evidence, and produce reviewable calculations and outputs. Replacing a lender's full model or loan origination system is not a prerequisite.

What if the uploaded documents disagree?

is being designed to retain multiple source observations, flag differences, and ask an analyst to accept or document the applicable value. It should never silently guess missing critical financial inputs.

Which real-estate asset classes are in 's scope?

The first scope is Indian sale-led residential development and construction finance. Stabilized office, retail, rental-property cash flows and international CRE standards are outside the initial release.

Is available for institutional deployment today?

is an in-development product direction. The interface and data shown here are conceptual. We are seeking conversations and controlled historical-case pilot partners to validate accuracy, workflow fit and security requirements before production claims.

BE A DESIGN PARTNER

Bring a real underwriting problem.
We'll bring the questions.

If you work in Indian real-estate lending, NBFC credit, private debt or project finance, we'd value a short discussion about how your team reviews developer models, source documents and credit risks.

Prefer a conversation? Message the founder on LinkedIn Exploratory discussion · No financial files needed

Join the pilot

Tell us a little about your team. We'll get in touch to explore a fit.

Your pilot signup details

* Required fields. Please don't include confidential project or financial data.