Bucketwise Join waitlist

Free credit-risk SQL prompt tools

Seven free tools for credit-risk analysts at banks and NBFCs. Type your table and column names, copy the prompt into your approved AI tool, then check the answer and present it. No customer data, ever.

Open the tools

Schema profile

Describe your tables once. Every other prompt works better when the AI knows your tables. Fill this in once and paste the result at the top of any prompt.

You enter: Table name, One row is..., Key column(s), Columns (one per line: name, type, meaning). You get: a ready prompt, 4 checks to run on the answer, what healthy and warning results look like, and how to present it.

Definition sheet

Lock your definitions. Two teams can compute different DPD from the same data. Write each definition once, then paste it into every prompt.

You enter: DPD counting, Buckets, NPA / default flag, Bad flag for scorecards, Months on book (MOB), Active loan.... You get: a ready prompt, 5 checks to run on the answer, what healthy and warning results look like, and how to present it.

Data dictionary

Plain-English meaning of every column. Turns your table list into a plain-English data dictionary. The AI works from names only, so it marks every guess and lists what to ask your data owner.

You enter: Your tables and columns, Type of lending, Who will read it. You get: a ready prompt, 5 checks to run on the answer, what healthy and warning results look like, and how to present it.

DPD and buckets

Compute DPD from schedule and payments. Builds a prompt that writes the DPD and bucket query from your instalment schedule and payment tables.

You enter: Instalment schedule table, Loan id column, Due date column, Due amount column, Payment table, Payment date column.... You get: a ready prompt, 5 checks to run on the answer, what healthy and warning results look like, and how to present it.

Roll rate

Bucket movement month to month. Shows how many loans in one bucket move to a worse, same or better bucket next month.

You enter: Monthly snapshot table, Loan id column, Month-end date column, Bucket column (or DPD column), Balance column, From month.... You get: a ready prompt, 5 checks to run on the answer, what healthy and warning results look like, and how to present it.

Bureau scrub

Find stale, duplicate and impossible bureau data. Checks a stored bureau extract for problems before you derive variables or build a model.

You enter: Bureau table, Customer id column, Application id column, Pull date column, Application date column, Score column.... You get: a ready prompt, 5 checks to run on the answer, what healthy and warning results look like, and how to present it.

Rejection analysis

Why applications are declined. Breaks declines down by reason, channel and score band so you can see what policy is doing.

You enter: Application table, Application id column, Decision column, Decision values, Reject reason column, Decision date column.... You get: a ready prompt, 5 checks to run on the answer, what healthy and warning results look like, and how to present it.

Paid tools are on the way

More than a hundred more tools are being tested with risk teams. Join the waitlist to hear when they open.

Join the waitlist