Skip to content

Draft a credit appraisal note

Turn financials and site-visit notes into a structured first draft of a credit note for review.

Recommended: BuilderWritingData

Best for

  • Credit officers in banks and NBFCs
  • MSME loan files with a standard note format
  • Analysts who spend hours on the first draft

What you need

  • An AI tool your bank has approved
  • Balance sheet and P&L figures, with borrower identity masked
  • Your site-visit notes
  • Your bank's credit note template

Starter prompt

Copy it, fill in the parts in square brackets, and paste it into the AI assistant your organisation allows.

Act as a junior credit analyst at an Indian [bank/NBFC]. Using only the data below, draft a credit appraisal note in this structure: business overview, three-year financial trend, key ratios (current ratio, debt-equity, DSCR) with the working shown, observations from the site visit, risks, and open questions for the borrower. Figures are in [lakh/crore]. Do not invent any number; if a ratio cannot be calculated from the data, write 'Insufficient data'. Borrower is referred to as 'the Applicant'. Financials: [paste] Site visit notes: [paste]

How to use it

  1. 1.Mask the file

  2. 2.Run the prompt

  3. 3.Recalculate the ratios

  4. 4.Edit into your template

  1. 1

    Mask the file

    Replace the borrower's name, PAN, GSTIN and account numbers with 'the Applicant'. Use only a tool your bank has approved for this data.

  2. 2

    Run the prompt

    Paste the masked figures and notes into the starter prompt. Ask it to show the working for every ratio.

  3. 3

    Recalculate the ratios

    Check each ratio in your own spreadsheet. Correct any figure the assistant got wrong before reading the narrative.

  4. 4

    Edit into your template

    Move the checked draft into your bank's template and add your own judgement and recommendation.

Watch out. Models make arithmetic slips and can fill gaps with plausible numbers. Mask all customer data, use approved tools only, recalculate every ratio, and keep the recommendation your own.

What you end up with

A checked first draft of the credit note, ready for your judgement.

Read why AI matters in banking & insurance

Related use cases