Free tools / Rejection analysis
Rejection analysis prompt for loan applications
Breaks declines down by reason, channel and score band so you can see what policy is doing. It is free, needs no sign-up, and works from table and column names alone.
What you enter
You type names only: table and column names, never customer data.
- Application table
- Application id column
- Decision column
- Decision values
- Reject reason column
- Decision date column
- Score column
- Channel column
- Product column
- From date
- To date
- Other cut columns
How to use it
- Pick a clean quarter first, then compare to the previous one.
- Look at reasons, then at channels, then at score bands.
- Write the three biggest findings in one line each, with a number.
Checks to run on the answer
AI can write SQL that looks right and is wrong. Run these checks before you trust the result.
- Approvals plus rejections plus excluded equals total applications.
- Each application has at most one primary reject reason, or the multi-reason rule is written.
- Reasons not in your policy list are counted and reviewed.
- Months with a policy change are flagged on the trend.
- The approval rate matches the number your business head already quotes, or you know why not.
Open the tool
You get the ready prompt, the checks above, what healthy and warning results look like, and how to present the result to an approver.
Open the free toolsRelated guides
Using AI for credit-risk work without sharing customer data
You can get useful credit-risk SQL and analysis from an AI tool by sharing only table and column names. How the method works and what to keep out of the chat.
Using AIHow to check AI-written credit-risk SQL before you trust it
AI tools write SQL that runs but may be wrong. Ten checks to run on any credit-risk query: row counts, duplicates, DPD edges, bucket totals and a hand trace.