Solution
For credit risk teams, portfolio managers, and private credit funds: every borrower submission read every period, with deterioration surfaced as cited, threshold-tested alerts.
The problem
Manual surveillance moves as fast as analysts can read submissions. Across a growing portfolio that means annual reviews — and risk that had four quarters to build.
A margin slipping two periods in a row, liquidity quietly tightening, an operating report filed late and thinner than the last one. The early signals sit in documents nobody had time to read.
Large accounts get watched; the long tail gets sampled. The loss that surprises the committee usually comes from a file nobody opened since origination.
The product, not a promise
How it works
Take in the financials, operating reports, and compliance documents borrowers submit each period.
Pull the metrics that matter — revenue, margins, leverage, liquidity — into structured, source-linked data.
Track each metric against prior periods, covenants, and policy thresholds.
Surface deteriorating trends and emerging risk conditions with the evidence attached.
Route flagged accounts to credit and risk teams for human judgment and documented action.
Who it's for
Credit risk analyst
Chief credit officer
Risk & compliance officer
Portfolio risk rarely announces itself. It shows up as a margin that slipped two quarters in a row, a liquidity position quietly tightening, an operating report filed late and thinner than the last one. Manual monitoring catches these signals only as fast as analysts can read submissions — which, across a growing portfolio, means annually and unevenly. Borrower & Portfolio Company Risk Monitoring reads every submission, every period, and surfaces what changed.
The agents ingest the documents borrowers and portfolio companies actually submit: financial statements, operating reports, compliance certificates, bank statements. From each package they extract the metrics that define financial health — revenue, margins, leverage, liquidity, cash flow — into structured data linked to its source pages. Each period’s figures are compared against prior periods, covenant levels, and your policy thresholds, so the trend is visible, not just the snapshot.
When something moves — performance deteriorating, a threshold approached, an anomaly between what the financials say and what the bank statements show — the account is flagged with the specific evidence attached. Risk teams open an alert and see the metric, the trend, and the exact source line behind it, rather than a score with no explanation.
A risk-monitoring signal is only actionable if you can defend it — to a credit committee deciding on a workout, or to a regulator asking why an account was downgraded when it was. Every alert here carries full lineage: the extracted values, the source documents, the thresholds applied. Every review decision is taken by a human and recorded in the audit trail. That turns portfolio surveillance from a sampling exercise into a consistent, portfolio-wide discipline — the same standard applied to every account, every period, with the paper trail built as it happens.
Objections, answered
Every alert shows the extracted values, the source document and page, and the threshold that tripped — an analyst verifies the signal in one click. No account is downgraded or actioned without a named human deciding.
Yes. Metrics are compared against each account's own covenant schedule and your policy thresholds, not a generic benchmark, and the comparison basis is recorded with every alert.
The full lineage: what was submitted, what was extracted, which thresholds were applied, what was flagged, and who decided what in response — recorded as it happened, per account, per period.
The platform reads the documents borrowers already submit, so there is no new borrower-facing process. Onboarding is loading covenant schedules and thresholds; monitoring starts with the next submission cycle.
Watch a quarter's submissions become cited, threshold-tested alerts live in the demo.
Request a demo