AI Threat Assessment · 26 May 2026

Tricog Health

AI Medical Diagnostics
STILL BREATHING
3.2/ 10

“Tricog built exactly what healthcare needed: AI that reads ECGs faster than cardiologists, with human specialists verifying every diagnosis to keep the lawyers happy. Ten years, 36 million patients diagnosed, regulatory approvals across multiple countries — this is the rare medtech company that shipped real product instead of PowerPoint promises. The twist is that the same foundation models now making their AI-plus-human workflow look quaint are about to make standalone AI diagnosis not just acceptable, but legally mandated.”

Business Model
5.0
Automation Risk
4.5
Moat Strength
2.5
Adaptability
3.0
Need Survival
2.0
AI Threat Level
BUSINESS MODEL REPLACEABILITY

Their pitch is 'AI + human verification = trust' — which worked beautifully when AI was unreliable and regulations demanded human oversight. ChatGPT-4V and Claude 3.5 Sonnet now read ECGs with cardiologist-level accuracy, and the FDA is fast-tracking standalone AI diagnostics that skip the human bottleneck entirely.

5.0
WORKFORCE AUTOMATION RISK

Those 24×7 cardiac specialists verifying every AI diagnosis? That's the exact job GPT-4V was trained to eliminate. The verification layer that justified their premium is becoming the cost center that kills their margins.

4.5
MOAT STRENGTH

Ten years of proprietary cardiac data, regulatory approvals in multiple countries, and 6000+ active users with workflow integrations create genuine switching costs. The medical device regulatory moat is real — getting FDA, CE, and CDSCO approvals takes years and deep clinical validation that Midjourney can't replicate by Tuesday.

2.5
AI ADAPTABILITY SIGNALS

Their careers page shows hiring for 'AI Engineers' and 'Machine Learning Scientists' — they're clearly rebuilding the engine while the plane flies. The question is whether they're upgrading to standalone AI fast enough, or just adding more verification layers to a workflow that verification is busy deleting.

3.0
WILL THE NEED SURVIVE AI?

Cardiac diagnosis absolutely survives — heart attacks don't pause for technology cycles. But 'AI-aided specialist-verified diagnosis' doesn't. The need shifts from 'verify what AI found' to 'trust what AI found,' and that transition happens faster than medical bureaucracy can adapt.

2.0
Verdict

“Foundation models will eliminate the human verification layer within 3 years, but the regulatory approvals, clinical integrations, and ten years of validation data buy them time to pivot to standalone AI before the workflow disappears entirely. They're the bridge between 'doctors don't trust AI' and 'doctors can't practice without AI' — and bridges get very rich during the crossing, even if nobody needs them on the other side.”

Scores are based on public information and AI analysis. This is an affectionate roast, not a financial assessment. The best companies use this as a mirror, not a verdict.

Roast another →