Central Uplift
cross-PMS data infrastructure for multi-site veterinary operators
See the demo →

Your hospitals are on different PMSes. Your reporting shouldn't be.

Central Uplift reads your hospitals' data out of ezyVet, Instinct, Cornerstone, AVImark, and the rest, and reconciles what was charted against what was billed across all of them — completed care that never made it onto an invoice, including the older sites that catch none of it. Below is a synthetic four-hospital network we built to show what that looks like.

Riverbend Veterinary Group is fictional. The data is grounded in AVMA and AAHA published research, with the AAHA Veterinary Fee Reference (11th ed.) for fees; constructed scenario rates are labeled as Central Uplift estimates. Methodology: /docs/synthetic-data-methodology.

See the demo →

Which services drive revenue, and which leak it

Revenue attribution tells you what bills well. Charge integrity tells you what got delivered and never billed. Both rolled up to the VMG/AAHA Chart of Accounts.

Services driving revenue
what OttoPilot answers[2]
  • Professional Services $5.07M
  • Surgical Services $4.89M
  • Diagnostic Imaging $3.53M
Services leaking revenue
what Central Uplift answers

Ten things you can find in the demo

Each line below is an operating question a multi-site veterinary COO actually has. The link drops you into the specific dashboard slice that answers it. Every number on every dashboard ties back to a SQL query you can copy out of the lower-right drawer.

Completed care got charted but never made it onto a bill.
Charge integrity →
Across all four PMSes, $1,692,174 of care was charted or performed in the window but never billed — net of the realization factor, the dollar that would actually have collected. It is worst where you can see it least: Riverbend Specialty Akron (legacy AVImark, no native charge capture) misses 34% of charted line value, the Cleveland ER 14%, against 10% network-wide. Your single-system tools catch their own system; this is the seam between them.
Headline revenue says we grew. Same-store says we didn't.
Network overview →
Year-over-year reads +8.0% on the deck against a calibrated prior-year baseline. Take out the hospital we acquired in March and same-store is +1.2%. The dashboard breaks both numbers apart so the board narrative is honest. (Prior-year is a synthetic calibration constant; the demo seed has no prior-year rows.)
15% of the referrals one of our own sites earned closed somewhere else.
Referral loop →
705 referrals in the window originated from Westwood Animal Hospital; 109 completed at an outside specialty hospital instead of one of ours. Priced at a single defended per-episode value ($119,900 across the gap, the build shown in the drawer), this is the referral-loop view no single-PMS tool computes — it joins a referral recorded at one site against the completing visit at another. Measured, not contacted.
Our DACVIM oncologist is at 71%. The other one is at 48%.
Clinicians · utilization →
Same hospital, same specialty, same patient pool. The fix is a scheduling-template default change; the cost of not fixing it is one specialist quitting and the other one being underbilled.
Urgent care quietly downcoded a chunk of February's exams.
Procedures · mix →
Columbus urgent-care visits got billed under the brief-exam code (UC-EXAM-BRIEF, $81) when the visit complexity matched the complex code (UC-EXAM-COMPLEX, $202), after a Cornerstone template default changed on 2026-02-01. Same revenue-integrity gap as a missed charge — a wrong code instead of a missing line — not a demand problem. The shortfall is back-billable; the forward gap is one template fix.
Friday late-night ER walkouts run 2.4× our baseline.
ER patterns →
Cleveland ER walkout rate is 16% in the Fri 21:00–01:00 window vs 6.6% the rest of the week. The heatmap by day-of-week × hour shows the cluster, with the no-DVM-on-shift weeks lining up. One DVM rotation onto the Friday late shift closes it.
Cleveland sends discharge reports back to rDVMs 50% of the time. Everyone else is near 95%.
rDVM health →
When discharge reports back to the referring vet lapse, the referring-vet relationship is the thing at risk (2025 AAHA Referral Guidelines codify the discharge summary as the expectation; no public compliance rate exists, so the gap is shown as a measured construct). The dashboard points at the named referral-coordinator role that closes the loop.
67 of our families keep pets at three or more of our hospitals.
Cross-network owners →
These households have patient records at 3+ sites under one owner identity — a specialty patient at one hospital, an ER or urgent-care patient at another. Cross-PMS owner-identity reconciliation makes the count possible; no single-PMS analytics product surfaces it.
Pittsburgh Tuesday wait is 61 minutes. Every other day it's about 22.
Capacity →
The Tuesday 11:00–12:00 check-in band is over-booked by a template default. Redistribute the block across Wednesday and Thursday and the wait goes back to normal — and the bail-rate on long-waiters comes back as recovered revenue.
Every number on this site ties to a SQL query you can read.
Audit log →
Per-connector P50 and P95 latency, success rate, total rows ingested. The Akron 2026-03-15 row is the integration moment for the acquired site. The live-sync feed replays a 12-second ingest end to end. Your CTO can audit every line.

What's the actual deal.

Honest about what we are not: we don't replace your PMS, we don't sell EHR. Read-only ingest. Mutual NDA. Synthetic data in this demo; real data in pilots. Casey Coco, Central Uplift LLC, Pittsburgh PA.