Contract varianceThe first thing we run on your record · selecting the first groups

Your carriers signed a fee schedule. Let's see where the payments don't match it.

Every PPO fee schedule promises an allowed amount per code. Your ERAs say what was actually paid. Nobody reconciles the two line by line, so on some codes they quietly stop matching. Vella lays the discrepancies out, by carrier and by code, with the contract line beside each one. Some will be leased‑network rates or plan downgrades; some won't. You decide which ones to raise with provider relations. It needs two things from you: your ERAs turned on, and the fee schedules you signed.

Run it on your record How it works

Advise modeVella reads and reports. Nothing is posted, appealed or changed unless you decide to.

Your recordYour carriers, your fee schedules, your remittances — no borrowed benchmarks, no industry average.

Priced, not freeA defined engagement with a price you see before we start. You keep every finding either way.

What it looks like

One carrier. Forty‑one codes. Six of them quietly short.

Below is a contract variance report shaped like the one you'd receive: a modelled practice, a real carrier's code set, and the arithmetic done the way we'd do it on yours.

Contracted versus adjudicated — Delta Dental PPO

Willow Park · trailing twelve months · 4,181 claim lines examined

6 of 41 codes short‑paid

CodeContractedActually paidPer claimAnnual

D0120Periodic oral evaluation$43.40$43.40—Correct

D1110Prophylaxis, adult$88.00$82.15−$5.85−$7,254

D2391Resin composite, one surface, posterior$178.00$164.20−$13.80−$6,707

D0274Bitewings, four films$72.00$72.00—Correct

D2740Crown, porcelain/ceramic$1,050.00$968.50−$81.50−$7,824

D2950Core buildup, including pins$198.00$172.40−$25.60−$9,856

D4341Periodontal scaling, per quadrant$242.00$221.75−$20.25−$8,343

D3330Endodontic therapy, molar$1,320.00$1,188.00−$132.00−$6,864

Codes examined41 · across 4,181 claim lines

Codes paid correctly35

Annual variance, this carrier alone−$46,848

Modelled data, illustrative of the format. Yours is computed from your own claim lines against the fee schedule in force on each date of service, and every figure carries the source rows behind it.

Check it yourself, right now

Pull three recent EOBs from this carrier.
Find the allowed amount on any code above.
Compare it to your signed fee schedule for the same code.

If they match, there's nothing to raise, and you'll know in ten minutes. If they don't, you've just found money that was always yours.

Why it happens, and why nobody catches it

Schedules get updated without notice. Codes get downgraded on adjudication. A contract gets loaded into the practice system with a typo in 2019 and nobody re‑reads it. None of that is detectable at the level of a single claim — it only appears when you compare thousands of them against the agreement, which is exactly the work nobody has time to do.

How it works

Thousands of lines, compared the way nobody has time to.

There is no scoping call and no consultant on site. A read‑only adapter copies your record, Vella reconciles it, and you get a report you can check against your own EOBs.

Illustrative. Matches in mint, short‑pays in peach. Each one carries the claim line, the remittance line and the contract line behind it.

01 · Connect

Read‑only, and you approve what moves

The Vella Connect adapter reads your practice‑management system with credentials you create and can revoke. You see the inventory before a byte is transmitted. Nothing is written back, ever.

02 · Reconcile

Every claim line against the schedule in force

Each paid line is matched to the contracted amount for that code, that carrier, on that date of service: thousands of comparisons, not a sample. Variance is only claimed where the contract and the remittance can both be produced.

03 · Report

A variance report with the receipts

By carrier, by code, per claim and annualised, with the source rows behind every figure. What you do next (reprocess, appeal, renegotiate) stays your decision. Vella advises; it doesn't post.

Designed to run while the practice is closed. Timings on a real record are ours to prove, not yours to assume. Our design target is an evening for a single practice and longer for a group; the first runs will tell us — and you — how close we are.

Once the record is whole

The same record answers the next questions too.

Contract variance is the door. With your record extracted and reconciled, the questions that used to take a consultant a quarter are queries, and each one comes with the way to verify it yourself.

What we'd look forWhat it's worthHow to check it yourself

Accepted treatment never scheduledDiagnosed, presented, agreed — and then it stopped. Anything no longer indicated or already done elsewhere is stripped out before counting.

$61,000per month, per location

Run your treatment‑plan report, filter to accepted with no future appointment, and compare the count to ours.

Recall cycles lost to frequency limitsPatients rebooked at exactly six months, whose plans require more than six months to have elapsed.

340patients a year

Count how many hygiene visits you rebooked at a six‑month interval, then check how many were denied.

Denials that were never resubmittedRejected once, worked never — most often because nobody was told the rejection existed.

22%of denied claims

Ask your clearinghouse for denied claims with no subsequent submission. The number usually surprises people.

Northgate — nothing worth fixingRunning at plan, capturing near what its market supports, with no meaningful recovery available.

$0and that's the finding

Nothing to check. We'd rather tell you a location is fine than manufacture something to justify the exercise.

Illustrative, modelled findings shaped like a first report, not results from a customer. Yours are computed from your own record, and every figure carries the source rows behind it.

The last row is the one that matters. Any analysis that finds money everywhere it looks is an analysis designed to find money. Some of your locations are fine, and you should hear that from us before you hear it from your own numbers.

The adapter

What exactly are you installing? A fair question.

We've just asked you to run software on the machine holding every patient record you have. Here is precisely what it does and, more usefully, what it is not permitted to do.

What it does

Connects to your practice‑management database using read‑only credentials that you create and can revoke.

Reads patients, appointments, clinical entries, treatment plans, claims, payments and documents.

Shows you a full inventory before a single byte is transmitted, so you approve what moves.

Encrypts everything in transit to your account, and keeps a log of exactly what was sent and when.

Runs only inside the hours you allow, so it never competes with a busy Monday.

What it never does

Write to your database. Not a flag, not a timestamp, not a status field. The credentials it holds cannot write.

Modify, delete or reorganise anything inside your practice‑management system.

Look at other machines, drives or shares on your network. It sees one database and one file path.

Transmit anything unencrypted, or to anywhere other than the account you authenticated.

Stay behind after you remove it. Uninstalling ends the connection immediately.

Signed binaryVerifiable publisher, so your IT team can confirm what they're approving.

BAA in place firstSigned before the adapter is issued, not after the data arrives.

PHI tokenisedIdentifiers are replaced before anything reaches a model.

Full audit logEvery read and every transfer recorded, and the log is yours.

And getting out is the same size as getting in. Uninstall the adapter and the connection stops that moment. Ask us to delete your copy and we confirm it in writing, with the date and what was removed. There is no retention clause and nothing to negotiate.

The full picture: Vella Connect →

Plainly

What we will and won't promise.

A proof engine has to start by being honest about where it is. Here is where we are.

We haven't run this on a customer's record yet.

You'd be among the first groups. That is why the terms are written the way they are, and why the report comes with receipts rather than assurances.

A price you see before we start.

This is a defined, priced engagement, not a free audit. A free audit is worth what you paid for it; a priced one has to be right.

Findings you can check without us.

Every figure carries the claim line, the remittance line and the contract line behind it. If you can't verify it against your own paperwork, it doesn't go in.

If there's nothing, we say so.

Some carriers pay to the penny. Some locations are fine. You'll hear that from us before you hear it from your own numbers, and you keep everything we produced getting there.

The worst outcome is that your carriers are paying you correctly.

That's the real downside. You connect the adapter, the record is reconciled, and the money you suspected was leaking isn't. You'd still be better off knowing, and you'd still keep everything we produced getting there.

Your record, extractedThe full history in open formats, yours to keep or take somewhere else entirely.

Every finding, with its receiptsSource rows behind each figure, and a way to verify each one without us.

An honest damage reportWhat your current system couldn't produce, what came back broken, and what was already missing before we arrived.

Run it on your record How the adapter works

Tell us which practice‑management system you're on and how many locations. Everything after that is a conversation about your fee schedules.

The scalable operating system that actually improves financial performance on its own.

Launching Q4, 2026

hello@canopy.dental

Platform

Meet Vella Sentinel What‑If™ engine Morning Clarity Platform Architecture

Solutions

Solo Practices For DSOs Group Practices Switching to Vella HIPAA & Security

Company

About & Founders Investors Get a plan for your tools

Resources

Model your practice Canopy University

Intelligent practice. Zero friction.