Monitor portfolio · PCT-007 · Sourced card
Re-check the compliance cert
Re-runs a borrower’s compliance certificate from the agreement’s own definitions and shows each line that moves the ratio.
Worked example: The certificate shows 4.90x, in compliance. Re-run from the definitions it is 5.40x against a 5.25x maximum: the non-recurring cap wasn’t applied, a $1.2m pricing action was counted as cost savings, and cash netting ignored the $5m cap. See the walk-through ↓
Today vs automated
- Open monthly packsAnalyst · manual
- Rebuild EBITDA in ExcelAnalyst · manual, re-keying
- Recompute each ratioPM · manual
- Email / meeting hand-offAnalyst → PM
- A compliance certificate arrivesTrigger: starts on its own
- Read the packsAutomated · Document store
- Apply the agreement's definitionsAutomated · LLM
- Flag gaps vs the certAI draft, human review · Monitoring model
- Whether a gap is a definitional judgment call or a real problemApprove / override · PM
AutomatedAI draft, human reviewHuman onlyTools: Inbox / portal, Document store, LLM, Monitoring model, Loan system
Estimated valueestimate
120–2,200 analyst hours a year
$18k–$550k of time, vs an assumed $20k–$100k tool cost
Net: −$82k to $530k a year net of tool cost; at the low end of the assumptions it does not cover the tool cost.
Assumptions
- 3–6 analyst hours per certificate today; 0.5–1.5 hours with an agent, including review
- 80–400 certificates a year for one team
- $150–250 per hour loaded cost
- Tool or build cost $20,000–100,000 a year (both)
Estimate, based on the assumptions shown; your numbers will differ. How we estimate
What else it can do for the team
- Track headroom trend quarter by quarter
- Flag add-backs that hit their caps
- Draft the query to the borrower on any gap
- Update the watchlist score
- Pre-fill the quarterly portfolio review page
Worked walk-throughKey catch: The certificate shows 4.90x, in compliance. Re-run from the definitions it is 5.40x against a 5.25x maximum: the non-recurring cap wasn’t applied, a $1.2m pricing action was counted as cost savings, and cash netting ignored the $5m cap.
1. The input
SAMPLE: fictional company, illustrative figures. Not a real borrower, deal or lender group.
Bayline Logistics, a fictional sponsor-backed regional freight carrier, sends its Q2 2026 compliance certificate. Inputs: the certificate with its cost-savings schedule, the monthly packs, and the agreement’s definitions.
“Consolidated EBITDA means Consolidated Net Income plus… (d) non-recurring or unusual charges, not to exceed 10% of Consolidated EBITDA for such period (calculated before giving effect to this clause (d)); (e) projected net cost savings from actions taken within 12 months, not to exceed 20%… For the avoidance of doubt, cost savings exclude revenue enhancements and pricing actions. Consolidated Total Net Debt: … less Unrestricted Cash not to exceed $5,000,000. §7.11 Maximum Total Net Leverage Ratio: 5.25 to 1.00 for the quarter ending June 30, 2026.”
Credit agreement excerpts (SAMPLE).
| Compliance certificate, Q2 2026 (LTM Jun 2026, SAMPLE) | $m |
|---|---|
| Reported EBITDA | 27.8 |
| Non-recurring charges | 2.9 |
| Run-rate cost savings (schedule: warehouse consolidation 1.5, route optimisation 0.9, fuel-surcharge repricing 1.2) | 3.6 |
| Pro forma EBITDA, carrier acquired May 2026 | 1.5 |
| Consolidated EBITDA | 35.8 |
| Total debt | 185.8 |
| Less cash | -10.4 |
| Total Net Leverage (max 5.25x) | 4.90x · In compliance |
| Monthly packs, reported EBITDA $m | Q3 25 | Q4 25 | Q1 26 | Q2 26 | LTM |
|---|---|---|---|---|---|
| Sum of monthly P&L | 7.4 | 7.0 | 6.5 | 6.0 | 26.9 |
2. What the AI produces
| Correction | Change | Leverage | Why |
|---|---|---|---|
| Per certificate | 4.90x | ||
| Reported EBITDA to monthly packs | +0.13x | 5.03x | Certificate LTM is 0.9 above the sum of the monthly P&L |
| Non-recurring cap (10%) | +0.03x | 5.06x | Cap is 2.7 on 26.9 |
| Remove fuel-surcharge repricing | +0.18x | 5.24x | A pricing action, which clause (e) excludes |
| Cash netting cap ($5.0m) | +0.16x | 5.40x | Certificate nets all $10.4m of cash |
Result: 5.40x against a 5.25x maximum, a breach by 0.15x. The certificate shows 4.90x. To get back to 5.25x would take about $0.9m more EBITDA or about $5.0m of net debt paydown, if the agreement allows an equity cure.
Draft note to the deal team: “Q2 certificate shows 4.90x; our recomputation from the definitions is 5.40x vs 5.25x. Four items, each cited to the clause. Recommend we confirm with the borrower and counsel before any response; do not send a notice yet.”
3. What a person still checks
- Ask the borrower to reconcile the certificate’s reported EBITDA to the monthly packs. An audit adjustment could explain part of the 0.9.
- Read the actual cost-savings clause: some agreements allow revenue synergies, and the answer flips the fuel-surcharge item.
- With counsel, check the cure rights, the cure deadline and whether a reservation-of-rights letter is needed. The tool never drafts a default notice.
- Decide with the lender group how to respond; a recomputed breach is a question for the borrower first.
4. Where it breaks
- The recomputation is only as good as the definitions loaded: amendments and side letters change caps.
- Monthly packs are unaudited and can be restated.
- Cost-savings schedules are often a single line; without the itemised schedule the agent can’t test eligibility.
How you get thereFour steps from prompting to an agentic workflow; most teams start at step 2.
Needs: an approved AI tool, a confidentiality policy.
Risk: unchecked answers, data pasted outside policy.
Needs: field definitions, a template, document parsing.
Risk: breaks silently when formats change.
Typical starting point
Needs: inbox/doc-store access, system write-back, logging.
Risk: acting on a misread; needs review gates.
The automated lane above
Needs: shared data model, permissions, audit trail, exception queue.
Risk: errors compound across steps; needs monitoring.
Solution optionsBuy (21 vendors listed), build or hybrid, with the trade-offs of each.
Orion (Private Credit AI), Lumonic, Allvue, Chronograph, Cobalt, 73 Strings, Canoe Intelligence, Finley, Alter Domus, Intain, Indus Valley Partners (IVP for Credit), Alphastream, CovenantIQ, Oxane Partners, Alkymi, Henon, Claira, Ellis, Goldslate, Raven (RavenRisk AI), Hypercore
Fastest start; you adapt to the product's data model and pricing.
An LLM API, a document parser, a vector or document store, your data warehouse, and a review queue.
Fits your definitions exactly; you own maintenance and controls.
Vendor extraction feeding your own checks, templates and systems.
Common middle path; two things to integrate and govern.
Listing is not endorsement.
What changes for the team3–6 h → 0.5–1.5 h per certificate (estimate); coverage goes from a sample of borrowers, quarterly to every borrower, every delivery.
How it works
Full title: Don't trust the compliance cert: re-run the covenant math from the monthly pack, automatically
“The cert says 4.9x. The agent's math from the same pack says 5.4x. Here is the line item.”
Scenario
SAMPLE: Fictional 'Bayline Logistics' sends its quarterly compliance certificate showing 4.9x against a 5.25x test. The monthly packs suggest EBITDA is falling.
What the human still decides
Whether a gap is a definitional judgment call or a real problem, and whether to call the sponsor.
Why it matters
The certificate is the borrower’s own math. Re-running it from the definitions is the lender’s check, yet only 16% of managers in PwC’s 2026 survey call AI-enabled monitoring a current priority.
What the record shows
Checked 2026-10-15.
Know which test you are tying out. Viavi's October 2025 Wells Fargo term loan has no financial maintenance covenant at all (Article VI runs §6.01–§6.09 without one); the live leverage test is the incurrence one, with unlimited restricted payments at pro forma Total Leverage ≤3.50x. The same computed-EBITDA problem shows up on the lender side: First Eagle's facility lenders cut the advance rate on recurring-revenue loans to 50% from 60% on 1 Oct 2026, so the borrowing-base headroom moved without the borrower changing.
Try it: prompt and data
Using the covenant definitions in sections {x}, recompute total leverage and fixed charge coverage from the attached monthly packs for the test period. Compare with the certificate. Show each input with its source and flag any difference above 0.1x.Data
Walk-through uses SAMPLE documents for a fictional borrower. In practice the inputs are your own confidential borrower documents; keep them inside an approved environment.
Who uses it, vendors and build
- Segments
- Direct lending
- Who uses it
- Portfolio manager, Associate
- Build
- Vendor + in-house
- In-house
- Extraction + a definitions-to-formula mapping per deal + unit tests. Estimate: 2–4 weeks. Each new deal needs its definitions mapped.
- Last verified
- 2026-10-15
Vendors that claim this capability
- Alkymi: financial-extraction, covenant-tests, loan-admin
- Finley: loan-admin, covenant-tests
- Hypercore: loan-admin, covenant-tests
- Ellis: financial-extraction, loan-admin
- Indus Valley Partners (IVP for Credit): financial-extraction, loan-admin
- 16 more vendors claim related capabilities. See all vendors for this stage →
Tags reflect each vendor's own public claims at the linked page (checked 2026-10-15). Not tested, ranked or endorsed.
Sources and related
Next step
Get new walk-throughs
Not investment advice. Names marked SAMPLE are fictional. Found an error? See the corrections log.