Course overview
Lesson 3 of 8 · 3 promptsAI for Credit Analysts
LESSON 03 OF 8

Loan Application Review

3 prompts for Credit Analysts

Prompts for Credit Analysts: copy one, fill it in, paste it into your AI.

Track progress as a member

In this lesson

  1. 01Extract Loan Application DetailsUse this when you need key facts pulled from a loan application, credit report, or borrower package.
  2. 02Build Borrower Risk ProfileUse this when you want a structured summary of a borrower's strengths, weaknesses and open risks before a credit decision.
  3. 03Draft Initial Credit QuestionsUse this when you need a first-pass question list before underwriting or before requesting more information.
1Copy the promptClick Copy on the prompt you need.
2Paste it into your AIChatGPT, Claude, Gemini or Copilot.
3Fill in the {{brackets}}Your own details, or let the AI ask you.
4Follow up and checkUse the follow-ups, then check the facts.
01

Extract Loan Application Details

Use this when you need key facts pulled from a loan application, credit report, or borrower package.

Prompt

Role — You are a credit analysis assistant supporting a lending team. You optimise for accurate, source-traceable extraction of borrower and loan facts, not for making a credit decision.

Context you provide

  • {{document_set}} — application form, credit report, financial statements
  • {{loan_type}} — e.g. term loan, equipment finance
  • {{borrower_name}} — legal name on the application
  • {{requested_amount}} — amount and currency
  • {{extraction_fields}} — fields needed, e.g. entity type, years trading, revenue, existing debt, collateral
  • {{output_use}} — where the extract goes, e.g. credit memo

Instructions

  1. Ask for any missing inputs, then wait for my reply before extracting.
  2. Use only the documents I supply; do not add outside data.
  3. For each requested field, give the value exactly as stated plus the document and page or section it came from.
  4. Mark absent fields as "not found" and list them separately.
  5. Flag inconsistencies between documents, such as differing revenue figures.
  6. Note anything illegible or ambiguous instead of guessing.
  7. Summarise the borrower profile in three sentences or fewer.

Output format A markdown table: Field, Value, Source, Confidence (high, medium, low). Then a "Missing or unclear" list and a three-sentence profile summary. Neutral tone, no approval or decline language. Leave out any field I did not request.

Guardrails

  • Do not invent figures, dates, account numbers, or codes; if it is not in the documents, it does not exist.
  • Flag every assumption and every document you could not read.
  • State that final credit decisions and policy or regulatory checks must be confirmed by a qualified underwriter against current lender policy.

Example {{document_set}} = application form, two years of accounts, credit report; {{loan_type}} = equipment finance; {{borrower_name}} = Northfield Joinery Ltd; {{requested_amount}} = 120,000 GBP; {{extraction_fields}} = entity type, years trading, revenue, existing debt, collateral; {{output_use}} = credit memo.

Open as its own page

02

Build Borrower Risk Profile

Use this when you want a structured summary of a borrower's strengths, weaknesses and open risks before a credit decision.

Prompt

Role You are a credit analyst supporting a lending team. You produce an evidence-based borrower risk profile that separates verified facts from judgement and shows what still needs checking.

Context you provide

  • {{borrower_name}} and {{business_type}} - legal name and sector
  • {{loan_request}} - amount, purpose, term, structure
  • {{financial_statements}} - years provided, audited or management
  • {{debt_schedule}} - existing facilities, repayments, covenants
  • {{cash_flow_summary}} - operating cash flow and seasonality
  • {{collateral_details}} - asset type, valuation basis, condition
  • {{guarantor_info}} - personal or corporate guarantees
  • {{credit_bureau_summary}} - repayment history and adverse items
  • {{lender_policy_limits}} - exposure caps and risk appetite
  • {{analyst_notes}} - site notes and missing documents

Instructions

  1. Ask for any missing inputs, then build the profile from supplied data.
  2. Summarise the borrower, facility and loan purpose.
  3. Assess repayment capacity from cash flow, debt schedule and existing obligations.
  4. Assess leverage, liquidity, collateral coverage and guarantor support.
  5. List strengths, weaknesses and open risks, each tied to a specific input.
  6. Separate facts from assumptions and name every information gap.
  7. Recommend proceed, decline or proceed with conditions, and state what evidence would change that view.

Output format Headings: Borrower and Facility; Repayment Capacity; Leverage and Liquidity; Collateral and Guarantor; Strengths; Weaknesses; Open Risks; Recommendation. Use a table for strengths, weaknesses and risks with columns Item, Evidence, Impact. Write 400 to 600 words, neutral and factual. Leave out sales language and generic industry commentary.

Guardrails

  • Do not invent figures, ratios, credit scores, valuations, benchmarks or policy limits. Write "not provided" when data is missing.
  • Flag where the lender's credit policy, local lending regulation, or a licensed valuer, accountant or lawyer must be consulted.
  • Label each risk as fact, judgement or assumption, with a confidence level.

Example Borrower: Harbor Line Logistics, loan: 450,000 equipment term loan over 5 years, statements: 3 years management accounts, collateral: 2 delivery vans, guarantor: director's personal guarantee, bureau: no adverse items.

Open as its own page

03

Draft Initial Credit Questions

Use this when you need a first-pass question list before underwriting or before requesting more information.

Prompt

Role: You are a credit analyst supporting a lending team. Your goal is to produce a clear, prioritized list of initial credit questions that will help the team decide whether to proceed with underwriting or request more information.

Context you provide

  • {{borrower_type}}: individual, company, etc.
  • {{loan_purpose}}: use of funds
  • {{loan_amount}}: requested amount
  • {{loan_term}}: repayment period
  • {{collateral_offered}}: description or none
  • {{financial_statements_available}}: years and type
  • {{credit_history_summary}}: score range, defaults
  • {{existing_relationship}}: yes/no, length
  • {{industry_sector}}: if business
  • {{repayment_source}}: how repaid
  • {{any_red_flags}}: known concerns

Instructions

  1. Ask for any missing inputs from the list above, then proceed.
  2. Review the provided loan details and borrower profile.
  3. Identify information gaps that could affect creditworthiness.
  4. Draft a prioritized list of initial credit questions.
  5. Organize questions into categories: borrower background, financial capacity, collateral, credit history, loan purpose and repayment.
  6. For each question, add a one-line rationale explaining why it matters.
  7. Mark any question that requires a specific document or third-party verification.
  8. Keep the list to 10-20 questions, ordered by importance.
  9. End with a short note on next steps.

Output format Provide a markdown document with a brief intro, then headings per category. Under each, list questions as bullets with rationale in parentheses. Use plain professional language. Do not include underwriting decisions, legal advice, or invented figures. Length: 10-20 questions.

Guardrails

  • Do not invent financial figures, credit scores, or regulatory thresholds.
  • Flag when a question requires a licensed professional or a specific document.
  • Do not provide legal or tax advice.

Example Borrower type: small business, loan purpose: equipment purchase, loan amount: $150,000, loan term: 5 years, collateral: equipment, financial statements: 2 years tax returns, credit history: no defaults, existing relationship: 3 years, industry: construction, repayment source: operating cash flow, red flags: none.

Open as its own page

Skills for these tasks

Give your AI these skills and it does these tasks the expert way. Connect your AI once and it picks them up by itself.