Credit View reads every document in one loan case — the sanction letter, application form, CAM, KYC, invoices, insurance, RTO forms, bank statements, bureau reports and more — extracts the parameters that matter, and shows where documents disagree, with the document, page and original text behind each value. It then runs the required-document checklist and the lending-policy rules, and exports an audit-ready workbook.
When to use it
- A disbursal file has thirty or forty documents and someone must confirm that the amount, tenure, rate, EMI and asset details agree across all of them.
- Half the pages are scans or photos, and keying them by hand takes a morning per file.
- Audit or quality-control teams need a sample of files checked completely, with evidence for every finding.
- Case data has to be loaded into a credit processing system in that system's own layout.
What it does
- Classifies each page into one of 46 document types and extracts up to 250 loan parameters in 17 groups — applicant and co-applicant, loan, pricing, charges, asset, insurance, credit bureau, banking, cash flow, disbursal, approval and more — using the Document Extractor.
- Compares the same parameter across documents, with rules for each kind of value (exact for money and counts, calendar-equivalent for dates, tolerant of spelling and spacing for names, addresses and identifiers), and lists mismatches by severity.
- For every extracted value, keeps the source document, page, page type, the original text and an extraction confidence, so each side of a mismatch can be opened and read.
- Checks the bundle against the required-document list for the case type (37 documents for a new commercial-vehicle retail case, each marked block, warn or info) and reports what is present and what is missing.
- Evaluates twelve lending-policy rules, including the LTV cap, minimum bureau scores for applicant and co-applicant, tenure against asset life, cash-flow cover for the EMI, insurance hypothecation and cover, bureau enquiry velocity, the NACH mandate against the EMI, applicant age and disbursal reconciliation, each shown with its threshold and the actual value.
- Extracts the tables that single values cannot capture: the repayment schedule, bank-statement transactions, bureau account history and existing loans.
- Builds a case summary, credit and risk profile, loan economics with reconciliation, asset and insurance details, the approval trail and rule-based callouts.
Inputs and outputs
- Accepted files
- .pdf, .xls, .xlsx, .png, .jpg, .jpeg — up to 200 MB each
- Files per case
- Up to 25
- Processing tier
- Lite, Medium or High, chosen before the run; the cost is estimated before and shown after
- Access
- Pro — granted on request by the EDEN team
- Exports
- A seven-sheet case workbook (summary, parameters, mismatches, checklist, credit and risk, loan economics, policy flags) and a CPA mapping workbook of 174 columns laid out for a credit processing system
- Tailored to you
- Your document types, parameters, checklist, policy rules and the mapping to your credit processing system
Example
Example — fictitious data
- Loan tenure: 60 months in every document — matched.
- Interest rate: 9.1% in the sanction letter, 9.4% in the CAM — flagged high, with both pages linked.
- Checklist: bank statement missing — noted; policy: cash-flow cover 3.26× against a 1.5× floor — pass.
What this does not do
- Values read from scanned pages are only as good as the scan; each value shows a confidence so you can decide what to check.
- A mismatch means the documents disagree. Deciding which value is right — and any credit decision — stays with you.
- The checklist and policy rules currently cover new commercial-vehicle retail cases; other case types run the same extraction and consistency checks.