Templates for the questions a lender asks before handing out money and after: which applications to approve, which live loans are drifting toward default and why, what losses the whole book carries, where it is dangerously concentrated, and whether asset quality is improving. Every template ends with a concrete result: a ranked list, a defensible decision rule, or a verdict with reasons in plain words. This section has 5 templates.
Every template is a live walkthrough on demo data: open it in the Templates section of Analytics Agent and press “Use this template” to run it on your own loan book.
Credit Scorecard (Approve Or Decline, Defensibly)
Shows: an application turns into points, a score, and a cutoff you can defend to an auditor; you see what actually predicts default, what each group of values costs, and what approving at each cutoff costs.
You get: an approve-or-decline decision made with a number and a justification on one page — something you can defend to an auditor instead of explaining by gut feel.
Default Classifier (Who Goes Bad, And Why The Model Says So)
Shows: the loan book ranked by default probability, three plain-language reasons attached to every row, and an alert threshold chosen by the cost of a missed default — not by the arbitrary 0.5 mark.
You get: bad loans show up in advance, and every decision can be explained in words — not as “the model said so”.
Expected Loss (What A Bad Year Costs)
Shows: the book’s annual losses not as one number but as a distribution: what a normal year costs, a year that happens once in twenty, and once in a hundred — with the reserve size for each.
You get: an honest price for a bad year built into the budget, and a separate reserve held for rare heavy years — fewer surprises at year end.
Credit Concentration (Where The Book Is One Bet)
Shows: a concentration index by industry and region, a “segment × rating” heatmap, and the ten largest borrowers; alerts fire on documented thresholds, not on gut feel.
You get: you can see where the book stands one step from a bad quarter — and what to do about it: diversify or set limits.
Bank Asset Quality (Better Or Worse)
Shows: the share of nonperforming loans and reserve coverage against a live corridor of comparable banks, capital context, and a summary verdict — whether the bank’s loan book is improving or falling apart — visible at a glance.
You get: instead of a bare number, a “better or worse” answer against market peers; deterioration shows up early, before the bank becomes a supervisory case.