An AI engine for commercial loan review.
Klerum runs the full lifecycle with AI doing the reading, extraction, sampling, and drafting: scoping, trial ingestion, sample recommendation, credit-file analysis, cash-flow normalization, findings, and memo issuance. Your reviewers stay in control of every decision.
AI Sampling
Upload a loan trial and the AI recommends a review population with per-borrower rationale. Refine it conversationally in plain English.
AI Extraction
The AI reads credit presentations, financial statements, tax returns, personal financial statements, and collateral valuations, returning key terms and figures structured and spread.
Key Considerations
The AI reads the whole relationship and surfaces the strengths, watch items, and risks a senior reviewer would raise, each with the questions to chase.
Cash Flow
Per-borrower normalization with live DSCR and surplus, plus global cash-flow groups that combine DSCR across related entities and guarantors.
Findings & Memos
An inline findings and exception workflow, AI memo drafting with cited evidence, and an independent AI peer review on every rating.
Loan Policy & Doc Review
The AI turns the bank's loan policy into testable standards and drafts the documents each loan should have on file, checking both and raising exceptions.
Methodology
Your standard scope as software. Reusable engagement profiles tell the AI how to sample, at firm or institution level.
M&A Due Diligence
Re-grade a target's loan book, mark the incremental credit reserve, and draft portfolio diligence takeaways from anonymized figures.
Customer Portal
A bank-side portal with edge-enforced access scoped to a single institution. Read-only engagement visibility today.
Fourteen tabs, one relationship workspace.
Each borrower relationship opens into a single workspace: Summary, Loans, Analysis, Financials, Cash Flow, Collateral, Global Cash Flow, Guarantors, Monitoring, Documents, Doc Review, Exceptions, Key Considerations, and Notes. Related entities and guarantors roll up into a global cash-flow group automatically, so combined DSCR is one click away, and every status change lands in the audit trail.
Modern AI, measured and transparent.
Klerum is built on Anthropic's Claude models and runs in production today. Reading a credit document costs a few cents in AI, measured, and that cost is passed through with a modest markup you can audit. The result is an AI platform a community bank can actually put to work.
See Klerum on your own portfolio.
A 30-minute demo, no deck. Upload a sample trial and watch the AI propose a review population against your methodology.