From policy wording to financial impact.
Know your cash-flow risks and protection gaps before a loss hits. SBFinance combines your financial export, an insurance-policy PDF and a short risk questionnaire to show where you may be financially exposed, what coverage may apply, and how an uncovered event could create a cash gap.
Built to help close the SME protection gap · All demo data is synthetic and clearly labelled
No install, no account, no upload · Deterministic calculations · Evidence from the numbers and the policy wording.
Most SMEs are underinsured — and can't see it
Small businesses rarely have the time, data or specialist knowledge to connect their own numbers to their policy wording. So they don't know where an uninsured event — a cyber outage, a key supplier failing, property loss, or an interruption — would actually hurt, or how it would hit cash flow.
Built for the growing SME
The illustrative case is a synthetic Singapore digital-services SME (SGD) with two sites, a dominant cloud-hosting supplier and rising digital dependency — the kind of Southeast-Asian business that is exposed to cyber, supply-chain and business-interruption risk but has no in-house risk team. As it grows, its protection needs change; SBFinance re-runs the numbers whenever the data does.
Three inputs you already have
Financial export
Upload a CSV/XLSX. SBFinance runs its deterministic analysis (revenue, expenses, cash, runway, supplier concentration).
Insurance policy PDF
Upload a text-based policy. SBFinance extracts limits, sub-limits, deductibles and exclusions with page-level evidence — scanned PDFs are safely unsupported.
Short risk questionnaire
A few fields (industry, digital dependency, key-supplier dependency, tolerable downtime, and any asset value you choose to enter).
Cyber, supply-chain, asset & business interruption
Cyber disruption
Cost of a selected outage = downtime × daily fixed-cost burden × activity affected + your recovery cost. No invented breach probabilities.
Supply-chain interruption
Spend-concentration signals plus an interruption scenario. We flag that high spend isn't automatically business-critical until you confirm it.
Asset / property loss
Uses the asset value you provide — never inferred from your expense transactions.
Business interruption
Days × daily revenue × contribution margin + unavoidable fixed costs, plus a user-set revenue-decline recovery window — shown against your cash buffer.
See exactly where coverage may fall short
An illustrative example (synthetic Singapore SME, SGD). Statuses stay honest — under ambiguity we show a range, not a false number.
| Risk | Status | Modelled impact | Limit | Uncovered |
|---|---|---|---|---|
| Cyber | No matching coverage found | SGD 44,444 | — | SGD 44,444 |
| Supply-chain | No matching coverage found | SGD 48,944 | — | SGD 48,944 |
| Business interruption | Possible partial gap | SGD 54,444 | SGD 25,000 (sub-limit) | SGD 30,444 |
| Asset / property | Possible partial gap | SGD 250,000 | SGD 200,000 | SGD 52,500 |
Indicative only. Actual coverage depends on the full policy wording, endorsements, exclusions and the insurer's decision. Not a coverage determination.
How a loss hits your runway
The scenario engine recomputes deterministically as you change downtime and the revenue-decline assumption — showing covered vs uncovered amounts and your runway and cash-gap date before and after the event. In the demo, a 30% revenue decline through recovery cuts the modelled runway from 12 months to about 6 and pulls the cash-gap date months earlier.
Every result shows its work
Policy facts link back to the page and the exact snippet they came from. Financial inputs show their derivations. The readiness score exposes its components, weights and missing-data penalty. Raw extracted statements, normalized facts and inferred interpretations are kept separate — an inference is never shown as a stated policy fact.
Runs on your machine
Ingestion, extraction, calculation and reporting run on-device in the desktop app — no cloud dependency. The uploaded policy is treated as untrusted input: SBFinance extracts data only and ignores any instructions embedded in the document. Policy identifiers are masked in the UI and logs, and derived exports can exclude raw transactions and policy identifiers entirely. (This public demo is a synthetic, read-only illustration and uploads nothing.)
What this is — and isn't
- Decision-support, not a broker, underwriter, claims handler or coverage determination.
- Text-based policy PDFs only — scanned/image PDFs are detected and marked unsupported (no OCR, no fabricated coverage).
- Loss figures are explicit user-selected "what-if" scenarios, not actuarial predictions or breach probabilities.
- Real-world extraction precision/recall is being measured on a first, synthetic benchmark; a broader real-policy corpus and SME/actuary review of the formulas are still to be done.
- No live insurer integration — a consent-gated, versioned JSON export API supports a future pilot.
- Cross-currency amounts are never combined without an explicit exchange rate and date that you provide.
Design-partner ready
A versioned Protection Assessment Export API produces a documented JSON export with explicit consent, redaction controls (derived risk facts only, or full) and an audit trail — so a future broker/insurer pilot can integrate the derived risk view without raw financial or policy data leaving the SME unless they choose. No insurer-specific endpoints are fabricated, and no affiliation with any insurer or programme is implied.
Frequently asked questions
Is this an insurance quote or a coverage decision?
No. It is indicative decision-support. Actual coverage depends on the full policy wording, endorsements, exclusions and the insurer's decision. Always verify with a qualified insurance professional.
Where does the analysis run?
Locally, on your machine, in the desktop app. There is no cloud dependency for the analysis, and the policy PDF never has to leave your device. This public browser demo is a synthetic, read-only illustration.
What files do you support?
CSV/XLSX financial exports and text-based insurance-policy PDFs. Scanned (image-only) PDFs are detected and safely marked unsupported — we never fabricate coverage from an unreadable document.
Do you estimate breach probabilities or asset values?
No. Loss scenarios are explicit "what if" calculations you control, and asset values are only ever the ones you enter — never inferred from your transactions.
Are the numbers reproducible?
Yes. The engine is deterministic and versioned — the same inputs always produce the same outputs. The browser demo is driven by a fixture generated from that same engine.
Important notices
Results are indicative and for decision-support only. SBFinance is not a broker, underwriter, legal adviser, claims handler or authoritative coverage engine. Policy interpretation requires human verification; actual coverage depends on the full policy wording, endorsements, exclusions and the insurer's decision. Estimated losses are scenarios, not actuarial predictions. SBFinance does not recommend or sell a specific insurance product. Consult a qualified insurance professional before purchasing, changing or relying on coverage. All examples on this page use synthetic data and an illustrative policy.
Run the Protection Gap demo
The interactive browser demo loads a synthetic SME, an illustrative SGD policy and a risk profile, then shows the full assessment — with sliders to change the disruption scenario. It runs entirely in your browser and is driven by a fixture generated by the same deterministic engine as the desktop app. For your own (private) data, install the local-first desktop app.