AI-Stroke

Name: AI-Stroke
Brand: AI-Stroke

AI Native

Building an AI neurologist to save the lives of stroke patients by providing better pre-CT diagnostics.

Est 2022|Valuation $44.1M|Raised $14.5M

LinkedIn 3K|Website|FTEs 7|HQ Occitanie, FR

Published Jul 14, 2026

01Common complaints

- Organizations that require an FDA-cleared, production-proven stroke AI solution for U.S. clinical use at the present time should consider incumbent vendors with documented regulatory clearance and broad deployment. - Workflows that are primarily imaging-centric and already optimized in-hospital (CT/CTA/CTP interpretation or interventional team activation) may achieve more direct benefit from established imaging AI platforms. - Operational environments that cannot support reliable video capture (for example, frequent low-light conditions, high noise, lack of patient cooperation, or other situational constraints) may not be suitable for a video-based assessment tool. - Procurements that require published SOC 2/ISO 27001 certifications and comprehensive security collateral prior to evaluation may find the vendor’s public documentation insufficient; expect to request additional security evidence.

Deal disqualifiers (likely)

Lack of regulatory clearance or appropriate intended use for the buyer’s jurisdiction
Absence of acceptable security and privacy posture for storing and transmitting patient video
No clear integration path into EMS documentation and hospital activation workflows

Publicly acknowledged limitations

No explicit limitations or formal contraindications were documented on the marketing pages accessed.

Practical limitations buyers should assume until proven otherwise

Model generalizability may vary across geographies, populations, languages, and EMS practices.
Clinical ground truth for stroke versus non-stroke can be complex without imaging and definitive diagnosis; label noise is a risk in training datasets.
Roadmap uncertainty: vendor references multi-site clinical studies and regulatory pathways, indicating product maturity is still evolving.

02Why they buy

Vendor-reported features and claims

The vendor describes the product as an AI-based clinical decision support tool for pre-hospital stroke diagnostics via mobile devices.
The vendor reports a dataset of 20,000+ stroke patient videos and 6 million images.
The vendor reports improved detection performance in a training-based comparison with over 2,000 EMS responders using real patient videos.

Assessment of differentiation

Pre-CT / pre-hospital video analysis represents a different decision point than imaging-based AI tools. This difference may be valuable if clinical effectiveness is demonstrated, but it also introduces higher operational risk in uncontrolled environments.

03Field reports

Review platform availability (Gartner Peer Insights, G2, TrustRadius, Capterra)

Searches did not find AI-Stroke listings or reviews on G2, TrustRadius, or Capterra that are accessible and indexed publicly.
Gartner Peer Insights: no public AI-Stroke review page was found in accessible search results; Gartner Peer Insights often requires login and may not index products publicly.

Implication: There is currently no robust corpus of third-party end-user reviews for AI-Stroke on mainstream B2B review platforms.

Community/Social (Reddit, Hacker News)

Searches for "AI-Stroke" on Reddit primarily returned general discussions about AI and stroke or unrelated references; searches did not surface substantive user feedback about the vendor website.
Searches for Hacker News discussions did not surface relevant threads specific to AI-Stroke.

Vendor-published testimonial-style claims

The vendor website includes a training-module comparison involving a large cohort of EMS/first responders (in collaboration with EMS centers), presenting examples such as: - Higher sensitivity (example shown: AI higher than EMS) - Higher overall accuracy (example shown: AI higher than EMS) - Similar specificity in the example (example shown: AI similar to EMS) Critical notes: - The described design appears to be a training context where responders labeled stroke signs from videos rather than a prospective clinical trial. - Ground truth methodology and patient selection criteria are not fully disclosed on the marketing page.

AI-Stroke case studies

No detailed, independently verifiable case studies (with named customer organizations, baseline metrics, and externally auditable outcomes) were found on the public AI-Stroke site pages accessed.

Adjacent market evidence (context, not AI-Stroke-specific)

Established imaging AI vendors such as Viz.ai publicly present outcomes and impact claims and have FDA-cleared algorithms and wider deployment, illustrating a more mature competitive position in imaging-based triage.

Critical takeaway: AI-Stroke appears earlier-stage relative to major imaging AI vendors. Buyers should expect pilots, careful monitoring, and staged rollout if proceeding.

04Best fit

Best-fit organizations

EMS agencies or fire/rescue services operating structured stroke protocols
Hospitals and regional stroke systems that require improved pre-notification and triage routing
Regions with long transport times where earlier suspicion could influence destination decisions

Organization size / scale

Likely best suited to mid-to-large EMS systems or hospital networks that can fund pilots, training, and governance.

Roles and personas

Decision makers: EMS medical directors, stroke program directors, ED leadership, CMIO/CNIO, regional stroke system coordinators
End users: paramedics / EMTs, triage nurses
Influencers: compliance/security officers, data protection officers, clinical quality teams

Pain points addressed

Missed or delayed stroke recognition in the field
Variability in clinician skill with subtle or atypical stroke presentations
Need to reduce time to treatment through earlier suspicion and triage

05The pitch

What the product does (core value proposition)

AI-Stroke is described by the vendor as an AI-based clinical decision support tool for pre-hospital / pre-CT stroke assessment and triage, delivered as a mobile app that runs on a smartphone or tablet. The core workflow described by the vendor is: - A first responder (or triage nurse) records a short patient video (the vendor cites a brief clip). - The AI analyzes facial symmetry, arm movement, and speech to detect signs consistent with stroke; the vendor states the analysis completes rapidly, prior to CT imaging. This is a distinct positioning compared to many established stroke AI vendors (e.g., Viz.ai, RapidAI), which primarily focus on in-hospital imaging workflows (CT/CTA/CTP) and care coordination rather than video-based pre-CT assessment. Vendor performance statements should be verified independently: AI-Stroke reports vendor-reported examples such as a claim of detecting a higher number of true stroke cases and claims of outperforming EMS responders in a training-based study. The vendor site provides limited methodological detail (dataset composition, ground-truth labeling, statistical significance, inclusion/exclusion criteria), so these claims should be treated as vendor-reported and not independently validated without peer-reviewed publication(s) or access to detailed study methodology.

Target market and market positioning

Primary users: EMS/paramedics, firefighters/first responders, and potentially triage nurses.
Primary buyers: EMS organizations, hospitals/health systems operating EMS or ED triage workflows, and regional/national emergency services.
Positioning: pre-hospital stroke diagnostics and pre-CT triage; vendor messaging emphasizes time-to-treatment outcomes.

Get a comprehensive analysis of AI-Stroke including market position, competitive landscape, adoption trends, and peer benchmarks.

Download full report →

Published Jul 14, 2026

01Common complaints

Organizations that require an FDA-cleared, production-proven stroke AI solution for U.S. clinical use at the present time should consider incumbent vendors with documented regulatory clearance and broad deployment.
Workflows that are primarily imaging-centric and already optimized in-hospital (CT/CTA/CTP interpretation or interventional team activation) may achieve more direct benefit from established imaging AI platforms.
Operational environments that cannot support reliable video capture (for example, frequent low-light conditions, high noise, lack of patient cooperation, or other situational constraints) may not be suitable for a video-based assessment tool.
Procurements that require published SOC 2/ISO 27001 certifications and comprehensive security collateral prior to evaluation may find the vendor’s public documentation insufficient; expect to request additional security evidence.

Deal disqualifiers (likely)

Lack of regulatory clearance or appropriate intended use for the buyer’s jurisdiction
Absence of acceptable security and privacy posture for storing and transmitting patient video
No clear integration path into EMS documentation and hospital activation workflows

Publicly acknowledged limitations

No explicit limitations or formal contraindications were documented on the marketing pages accessed.

Practical limitations buyers should assume until proven otherwise

Model generalizability may vary across geographies, populations, languages, and EMS practices.
Clinical ground truth for stroke versus non-stroke can be complex without imaging and definitive diagnosis; label noise is a risk in training datasets.
Roadmap uncertainty: vendor references multi-site clinical studies and regulatory pathways, indicating product maturity is still evolving.

02Why they buy

Vendor-reported features and claims

The vendor describes the product as an AI-based clinical decision support tool for pre-hospital stroke diagnostics via mobile devices.
The vendor reports a dataset of 20,000+ stroke patient videos and 6 million images.
The vendor reports improved detection performance in a training-based comparison with over 2,000 EMS responders using real patient videos.

Assessment of differentiation

Pre-CT / pre-hospital video analysis represents a different decision point than imaging-based AI tools. This difference may be valuable if clinical effectiveness is demonstrated, but it also introduces higher operational risk in uncontrolled environments.

03Field reports

Review platform availability (Gartner Peer Insights, G2, TrustRadius, Capterra)

Searches did not find AI-Stroke listings or reviews on G2, TrustRadius, or Capterra that are accessible and indexed publicly.
Gartner Peer Insights: no public AI-Stroke review page was found in accessible search results; Gartner Peer Insights often requires login and may not index products publicly.

Implication: There is currently no robust corpus of third-party end-user reviews for AI-Stroke on mainstream B2B review platforms.

Community/Social (Reddit, Hacker News)

Searches for "AI-Stroke" on Reddit primarily returned general discussions about AI and stroke or unrelated references; searches did not surface substantive user feedback about the vendor website.
Searches for Hacker News discussions did not surface relevant threads specific to AI-Stroke.

Vendor-published testimonial-style claims

The vendor website includes a training-module comparison involving a large cohort of EMS/first responders (in collaboration with EMS centers), presenting examples such as:

Higher sensitivity (example shown: AI higher than EMS)
Higher overall accuracy (example shown: AI higher than EMS)
Similar specificity in the example (example shown: AI similar to EMS)

Critical notes:

The described design appears to be a training context where responders labeled stroke signs from videos rather than a prospective clinical trial.
Ground truth methodology and patient selection criteria are not fully disclosed on the marketing page.

AI-Stroke case studies

No detailed, independently verifiable case studies (with named customer organizations, baseline metrics, and externally auditable outcomes) were found on the public AI-Stroke site pages accessed.

Adjacent market evidence (context, not AI-Stroke-specific)

Established imaging AI vendors such as Viz.ai publicly present outcomes and impact claims and have FDA-cleared algorithms and wider deployment, illustrating a more mature competitive position in imaging-based triage.

Critical takeaway: AI-Stroke appears earlier-stage relative to major imaging AI vendors. Buyers should expect pilots, careful monitoring, and staged rollout if proceeding.

04Best fit

Best-fit organizations

EMS agencies or fire/rescue services operating structured stroke protocols
Hospitals and regional stroke systems that require improved pre-notification and triage routing
Regions with long transport times where earlier suspicion could influence destination decisions

Organization size / scale

Likely best suited to mid-to-large EMS systems or hospital networks that can fund pilots, training, and governance.

Roles and personas

Decision makers: EMS medical directors, stroke program directors, ED leadership, CMIO/CNIO, regional stroke system coordinators
End users: paramedics / EMTs, triage nurses
Influencers: compliance/security officers, data protection officers, clinical quality teams

Pain points addressed

Missed or delayed stroke recognition in the field
Variability in clinician skill with subtle or atypical stroke presentations
Need to reduce time to treatment through earlier suspicion and triage

05The pitch

What the product does (core value proposition)

A first responder (or triage nurse) records a short patient video (the vendor cites a brief clip).
The AI analyzes facial symmetry, arm movement, and speech to detect signs consistent with stroke; the vendor states the analysis completes rapidly, prior to CT imaging.

This is a distinct positioning compared to many established stroke AI vendors (e.g., Viz.ai, RapidAI), which primarily focus on in-hospital imaging workflows (CT/CTA/CTP) and care coordination rather than video-based pre-CT assessment.

Vendor performance statements should be verified independently: AI-Stroke reports vendor-reported examples such as a claim of detecting a higher number of true stroke cases and claims of outperforming EMS responders in a training-based study. The vendor site provides limited methodological detail (dataset composition, ground-truth labeling, statistical significance, inclusion/exclusion criteria), so these claims should be treated as vendor-reported and not independently validated without peer-reviewed publication(s) or access to detailed study methodology.

Target market and market positioning

Primary users: EMS/paramedics, firefighters/first responders, and potentially triage nurses.
Primary buyers: EMS organizations, hospitals/health systems operating EMS or ED triage workflows, and regional/national emergency services.
Positioning: pre-hospital stroke diagnostics and pre-CT triage; vendor messaging emphasizes time-to-treatment outcomes.

Get a comprehensive analysis of AI-Stroke including market position, competitive landscape, adoption trends, and peer benchmarks.

Download full report →