CAP Today on Clinlab.AI: What Happens When a Lab Finally Uses Its Own Data
CAP Today’s October issue profiles Marie, Stanley, and Elara, the three AI agents that run quality, operations, and billing for clinical labs. Here is what the story covers and what it means for your lab.
Every clinical lab produces thousands of patient results a week. Most of that data gets reported, filed, and never looked at again. In the October 2026 issue of CAP Today, our CEO Abdul Hamid Halabi named the problem directly: labs are “sitting on some of the richest data in medicine,” and they use very little of it.
The piece, written by Josh Baxt for CAP Today’s Newsbytes column, explains how the Clinlab.AI agentic AI platform puts that data to work in three places where labs lose the most: billing, quality control, and operations. CAP Today is published by the College of American Pathologists and read by pathologists and lab leaders nationwide. We’re proud to be in it.
Three Agents, Built by Lab Experts
Clinlab.AI was founded by Abdul Hamid Halabi, an engineer with years of experience applying AI in health care, and his wife, pathologist Gazal Alsaati, MD. That pairing shaped the platform. Each agent was designed around a specific job that labs struggle to do well with limited staff, and each carries a name with a story behind it.
Elara: Get Paid for the Work You Already Did
For most labs, the fastest money to recover is money they have already earned. Elara goes after it. It reads payer contracts to understand how reimbursement really works on the back end, across public payers, private payers, and every variation in between.
From there, Elara scrubs claims for coding errors before they go out, automates eligible corrections, and tracks reimbursement trends by payer and CPT code. It also shows which physicians drive denials and why, so billing problems become coaching opportunities instead of recurring losses.
The results at the first lab to use Elara: initial denials fell from 30 percent to 5 percent, final denials fell from 15 percent to 1 percent, and accounts receivable turnaround dropped by more than 80 percent.
A clean claim still depends on a result you can stand behind. That is Marie’s job.
Marie: Quality Control on the Quality Control
Marie, named for Marie Curie, monitors results using more than 20 QC algorithms alongside Westgard rules, Six Sigma logic, and real-time patient-based QC. Most labs run daily QC and review Levey-Jennings charts monthly. Marie runs those checks every day.
The bigger difference is that Marie learns your lab: your instruments, your reagent lots, your ordering physicians, and the results your patients typically produce. When something doesn’t fit that pattern, Marie flags it. That same knowledge lets it build reference ranges for your actual patients instead of a national average.
The article’s strongest example: in 2024, a lab’s vitamin D results shifted sharply even though calibration and QC both passed. The calibrator and the QC material came from the same manufacturer, so the error was invisible to standard checks. Marie caught it because the population hadn’t changed, but the share of patients flagged as deficient jumped from about 30 percent to 50 percent in a single week. No one else noticed for weeks. The finding contributed to the calibrator being replaced.
Passing QC is not proof your patient results are right. Marie checks both.
Stanley: Buy What You Will Actually Use
Where Marie learns your patients, Stanley learns the physicians who order from your lab: their schedules, vacations, volumes, and test mix. Named for Frederick Stanley Nowlan Jr., who helped develop reliability-centered maintenance, Stanley pairs that physician model with what Marie knows about your test menu and instruments. The result is an accurate forecast of daily volume and reagent needs, automated purchasing, and better staffing decisions.
Lab teams spend about an hour, once or twice a month, scanning inventory. Stanley handles the rest. Labs using it report an average 35 percent cut in inventory spend, no reagent outages, and far less waste.
It Works with the Systems You Already Have
All three agents run as one platform on top of your LIS using HL7, and connect to EMRs through HL7 or FHIR. Lab managers work with them through dashboards or email alerts. No rip-and-replace.
Why This Matters for Independent Labs
Abdul closed the interview with the point that drives everything we build: independent labs have limited resources, and they still have to get paid, deliver quality, and control supply costs. Most software helps with one of those. Our platform covers all three, and the agents share what they learn with each other.
That is also why the platform sits at the core of the Smart Labs we design, build, and operate for physician practices and health systems, including our recent partnership with MyCardiologist in South Florida, a CVAUSA practice.
Read the Full Story
Want to see what Marie, Stanley, and Elara would find in your lab? Explore the platform and book a demo.
Comments