The frameworks from the writing, made usable. Each one runs entirely in your browser — nothing is uploaded, stored or sent anywhere.
How a credit score is actually built. Move the five weighted factors — payment history, amounts owed, length, new credit, mix — and watch where the points go, which factor is costing you most, and what would genuinely move it.
Open explainer →
PAYDEX, the Delinquency Predictor (formerly Commercial Credit Score), the Failure Score and the D&B Rating — four measures answering four different questions. Enter trade lines and watch the dollar-weighting decide your PAYDEX.
Open explainer →
A fitted scorecard, not a mock-up. Ten bureau attributes, weight-of-evidence binned and run through a logistic regression, scaled to points. Every factor has a “not available” option — because missing data is a value, and this model prices it.
Open scorecard →
Build a lending rate from its components — cost of funds, expected loss, capital charge, operating cost — and test whether it clears your hurdle. Includes a break-even rate solver and four presets from MSME working capital to anchor-backed TReDS.
Open calculator →
Model the cash conversion cycle — DIO, DSO, DPO — and see what a change in collection performance actually frees up in rupees. Separates the DSO you granted from the DSO you're suffering, with Best Possible DSO, Average Days Delinquent and the CEI benchmark.
Open calculator →
An ECL provision matrix for trade receivables — a historical loss rate for every ageing bucket, a forward-looking macro overlay on top, and the Stage 1 / Stage 2 / Stage 3 split that sets your loss horizon. Four scenario presets from a benign cycle to a downturn.
Open calculator →
Run the four limit-setting methods side by side — requirement-based, capacity-based, insurer-led and portfolio concentration — and take the lowest, as you should. The binding method is highlighted, with the gap to the next-loosest shown as your growth headroom.
Open calculator →Most credit frameworks are published as prose and consumed as vibes. You read that expected loss is PD × LGD × EAD, nod, and go back to pricing off a grid. The arithmetic only changes behaviour when you can move a slider and watch the verdict flip.
So each calculator here pairs with a written framework. The writing explains why the components matter; the tool lets you find the point where your own numbers stop working. In my experience that second step is where the actual learning happens — particularly the moment a well-liked deal turns out to sit below the break-even rate.
Everything runs client-side in plain JavaScript. No accounts, no analytics on your inputs, nothing leaves your browser — which matters, because these get used with real portfolio numbers. All models are illustrative and simplify tenor, tax, transfer pricing and cross-sell economics. They are teaching tools, not credit decisioning systems.
If there's a framework you'd like turned into a calculator, tell me.