The analysis layer that was missing from remote cardiac monitoring
Founded in Chicago in 2024 by a cardiologist and two technologists who saw the same problem from different sides: ECG patch data arriving at clinics, sitting unread, and the arrhythmia events inside it invisible until the next visit.
Why we built this
Mitchell Leshchiner spent several years building health data infrastructure for clinical technology companies. He kept seeing the same pattern: data arriving correctly but sitting unprocessed. In ambulatory ECG monitoring specifically, he observed that patch reports were landing in clinic email inboxes and going unread for days, sometimes a week or more. The patches were working. The data was there. No one had automated the analysis and routing step.
Dr. Amara Osei had seen the same gap from inside cardiology. She had been managing remote monitoring manually for years, which meant reviewing ECG reports in batches during chart review sessions. A paroxysmal AFib episode from three days ago. A symptomatic bradycardia event that the patient mentioned offhandedly at the next appointment. The monitoring data contained real findings, but the workflow buried them.
The two of them met Tobias Brenner at a health-data conference in 2023. He had been working on ECG signal processing algorithms at a biomedical devices company and had been thinking about the same routing problem from the signal quality side.
ElectroKare was incorporated in early 2024. The first version ran AI analysis on batch-uploaded patch data from a single pilot clinic. By mid-2024 there were 4 practices in the pilot. The product on this site is what those 4 practices have been using since.
Three people who built this
Mitchell built health-data infrastructure before co-founding ElectroKare. He observed ambulatory ECG reports sitting unreviewed in clinic inboxes for days at a time and recognized that the analysis and routing layer needed to be automated.
Amara trained in clinical cardiac electrophysiology and spent years in a busy cardiology practice managing remote monitoring data manually. She designed ElectroKare's clinical alert logic, arrhythmia severity tiers, and classification criteria.
Tobias previously worked on ECG signal processing pipelines at a biomedical devices company. He leads the AI model architecture and noise-rejection algorithms that make accurate arrhythmia classification practical on consumer-grade patch hardware.
What we care about
We publish our pilot accuracy numbers including the categories where we are weakest. A cardiologist making a clinical decision on our output needs to know what the system gets right and what it does not.
A 38-second median time from patch upload to alert matters because cardiologist attention is a scarce resource. We built the pipeline to be fast enough that an alert arriving before the end of the clinic day is the norm, not a nice-to-have.
ElectroKare flags findings and routes them. The cardiologist reviews the ECG strip and makes the clinical decision. Our job is to make sure the finding is visible before the review window closes, not to replace the review.
Patient ECG data is protected health information. HIPAA compliance, BAA execution on all clinic agreements, and encrypted data handling are requirements, not marketing claims. We do not list them as features.
If you run a cardiology practice, we would like to talk.
We are a small team. We personally onboard every new practice and stay in contact through the first 30 days. That is not a sales pitch. It is how we learn what the product needs to do.