Viz.ai, Inc.
AI-powered disease detection and care-coordination company founded in 2016 whose Viz LVO stroke-triage software won the first FDA De Novo clearance for AI stroke triage in February 2018 and reached a $1.2 billion valuation in 2022.
Ask AI to summarize this page
Use this page as a primary source and ask the AI to separate verified facts from items that could not be verified.
FactsLedger cannot guarantee that an AI response will include a citation.
- Established
- 2016
- Location
- San Francisco, California
- Country
- United States
- Founded
- 2016
- Founders
- Dr. Chris Mansi (CEO); Dr. David Golan (CTO)
- Headquarters
- San Francisco, California, United States
- What it does
- AI-powered disease detection and intelligent care coordination, beginning with stroke triage
- First FDA De Novo
- February 2018 (DEN170073, device name ContaCT)
- Valuation
- $1.2 billion (Series D, April 2022)
Overview
Viz.ai is an artificial-intelligence healthcare company whose software detects suspected strokes in medical scans and alerts specialists within minutes (MedCity News). Founded in 2016 by Dr. Chris Mansi, a former NHS neurosurgeon, and David Golan, a machine-learning postdoc at Stanford, the San Francisco-based company built its name on a single regulatory first: in February 2018 its stroke-triage software became the first AI algorithm to receive FDA De Novo clearance for detecting large vessel occlusion strokes in CT angiography, creating an entirely new regulatory classification for computer-aided triage software (MedCity News; BioCentury). Four years later, a $100 million Series D led by Tiger Global and Insight Partners valued the company at $1.2 billion, with total venture investment above $250 million (MobiHealthNews).
The regulatory first: a De Novo for AI triage
Before Viz.ai, no AI triage device had been through FDA’s De Novo premarket review pathway. BioCentury reported that the FDA’s February 2018 approval of the Viz LVO Stroke Platform — granted via De Novo — opened up a new regulatory classification under the 510(k) pathway for future computer-aided triage software devices with the same intended use (BioCentury). The FDA’s own decision summary, DEN170073, records the device under the proprietary name ContaCT: a Class II device under special controls (21 CFR 892.2080, product code QAS) that analyzes CT angiogram images of the brain in the acute setting and notifies a neurovascular specialist when it identifies a suspected large vessel occlusion (U.S. Food and Drug Administration). The indication is carefully bounded — the device is a “notification-only, parallel workflow tool,” and the suspected findings are not for diagnostic use beyond notification (U.S. Food and Drug Administration).
The submission data, as reported by BioCentury, came from a 300-patient retrospective reader and performance study comparing the software’s image analysis against standard of care: the software identified large vessel occlusions with 90% accuracy, the median scan-to-notification time was under six minutes, and it notified specialists faster in more than 95% of cases, saving an average of 52 minutes (BioCentury). The FDA’s decision summary independently records a standalone-performance area under the curve (AUC) of 0.91 for the algorithm (U.S. Food and Drug Administration).
The regulatory story did not end in 2018. In February 2022 the company landed an FDA 510(k) clearance for its algorithm designed to detect cerebral aneurysms, and its LVO software earned European CE Mark approval in 2021 (MobiHealthNews; Endovascular Today). Perhaps more consequentially for hospital economics, its AI software received approval for a New Technology Add-on Payment from the Centers for Medicare & Medicaid Services — the first such reimbursement for AI software (Endovascular Today).
The clinical evidence: does it change treatment times?
The profile of record for Viz.ai’s effectiveness is not its funding history but a peer-reviewed retrospective cohort study published in Internal and Emergency Medicine in 2025 (Springer). Researchers at a comprehensive stroke center analyzed 222 patients who underwent endovascular thrombectomy between July 2018 and July 2022 — 84 before the hospital implemented Viz LVO in July 2020 and 138 after — with thrombectomy within 90 minutes as the primary outcome (Springer). The results were substantial: median door-to-groin time fell from 109 minutes to 75 minutes (p < 0.001), and the fraction of patients receiving thrombectomy within 90 minutes rose from 26.5% to 68.1% (p < 0.001) (Springer). An interrupted time-series analysis confirmed the reduction was tied to the implementation itself (p = 0.001), not a background trend (Springer).
The authors were careful about what the study does not show: there were no significant differences in reperfusion scores, length of stay, or mortality and hospice rates, the study was not powered for those outcomes, and 90-day functional outcomes were unknown — a limitation of its retrospective design (Springer). It was also a single-center study, though its findings align with other groups’ reported door-to-groin reductions of 12 to 22.5 minutes after AI implementation (Springer).
A large real-world multi-center evaluation reported in Endovascular Today points the same way at scale: across 2,544 consecutive patients at 139 hospitals using scanners from multiple manufacturers, the Viz LVO workflow software showed a median time-to-notification of 5 minutes and 45 seconds, with 96% sensitivity and 94% specificity in identifying large vessel occlusions (Endovascular Today).
From stroke-only to a multi-disease platform
Viz.ai’s commercial arc is one of expansion from a single stroke product into adjacent time-critical conditions. In March 2021 it raised $71 million in Series C funding led by Scale Venture Partners and Insight Partners — bringing total investment above $150 million — explicitly to carry its platform into cardiology, pulmonary and trauma care (MedCity News). By April 2022 it had launched AI detection products for aortic disease, pulmonary embolism and cerebral aneurysms, with a 510(k) submission pending for subdural hematoma, and had begun building an AI-driven life-sciences platform to help medical-device and pharmaceutical companies run clinical trials (MobiHealthNews; Endovascular Today). In 2019, Medtronic struck a deal to distribute the company’s detection and triage software, widening its hospital distribution through the medtech incumbent (MedCity News).
The clearest sign of its move up the hospital technology stack came in November 2024, when Viz.ai entered an agreement with Microsoft to launch 13 of its AI workflow models on Microsoft’s Precision Imaging Network, part of Microsoft Cloud for Healthcare (AuntMinnie.com). The models — covering stroke, hemorrhage, pulmonary embolism and aortic dissection — are designed to help radiologists create reports more efficiently, streamline communication with treating physicians, and support care-team collaboration on follow-up treatment (AuntMinnie.com).
What this profile does not claim
This profile does not claim that Viz.ai’s software improves long-term patient outcomes — the peer-reviewed study found faster treatment times but no significant difference in reperfusion scores, length of stay, or mortality, and 90-day functional outcomes were not measured (Springer). It does not state post-2022 hospital counts, clearance totals, profitability, or partnership rosters, which appear only in company-issued material not read or verified here. It does not report seed, Series A, Series B, or Series E financing details, which were not verified from sources read in full. The FDA De Novo device is named “ContaCT” in the agency’s decision summary while press coverage calls the product the “Viz LVO Stroke Platform”; both refer to the same DEN170073 clearance (U.S. Food and Drug Administration; BioCentury).
Timeline
- 2016
- Viz.ai is founded by Dr. Chris Mansi, a former NHS neurosurgeon, and David Golan, a machine-learning postdoc at Stanford. The San Francisco-based startup sets out to build clinical decision-support tools for stroke detection.
- 2018
- The FDA grants De Novo clearance (DEN170073) for the company's Viz LVO / ContaCT stroke-triage software — the first AI algorithm to receive De Novo authorization for stroke detection in CT angiography, creating a new regulatory classification for computer-aided triage software. The submission rested on a 300-patient retrospective study showing 90% accuracy and a median scan-to-notification time under six minutes.
- 2019
- Medtronic strikes a deal with Viz.ai to distribute the company's detection and triage software, extending its hospital reach through the medtech incumbent's sales footprint.
- 2021
- Viz.ai raises a $71 million Series C led by Scale Venture Partners and Insight Partners (total raised above $150 million) to expand beyond stroke into cardiology, pulmonary and trauma care; its LVO software also earns European CE Mark approval.
- 2022
- A $100 million Series D led by Tiger Global and Insight Partners values the company at $1.2 billion, with more than $250 million raised to date; in February the FDA grants a 510(k) clearance for its cerebral-aneurysm detection algorithm, and its stroke software receives a New Technology Add-on Payment (NTAP) from the Centers for Medicare & Medicaid Services — the first such reimbursement for AI software.
- 2024
- Viz.ai enters an agreement with Microsoft to launch 13 of its AI workflow models on Microsoft's Precision Imaging Network, part of Microsoft Cloud for Healthcare, embedding its triage algorithms in clinical radiology workflows.
What we could not verify
These claims came up in research and could not be confirmed. They are recorded here rather than published as fact or quietly dropped. If you hold a source that settles one, please tell us.
- Post-2022 scale claims (hospital counts, FDA-clearance totals, profitability, life-sciences partnerships) appear only in company-issued anniversary material and are deliberately omitted.
- Seed, Series A, Series B and Series E financing details were not verified from sources read in full and are omitted.
- The DOTmed summary of the VALIDATE multi-center study could not be fetched (source-server failure); the clinical-evidence section uses the peer-reviewed Springer study instead.
- Exact current office locations beyond San Francisco (Tel Aviv, Amsterdam, Portugal) were reported only in sources not read in full; the profile states only San Francisco.
Sources
8 sources (2 tier 1, 6 tier 2), each opened and read against the claim it supports. Every source is independent of the subject; no first-party material is cited. See the source policy.
No first-party material was used. Every substantive claim rests on sources read in full: BioCentury and Endovascular Today (specialist medtech press), MedCity News and MobiHealthNews (health technology press), AuntMinnie (radiology trade press), the FDA's published De Novo decision summary DEN170073 (government record), and a peer-reviewed retrospective cohort study in Internal and Emergency Medicine. Company-issued press releases and wire republications of them were used only as leads for dates and were never cited.
- Supports: De Novo number DEN170073; applicant Viz.AI; device proprietary name ContaCT; Class II special controls under 21 CFR 892.2080, product code QAS; the device as a notification-only, parallel-workflow tool analyzing CT angiogram images of the brain in the acute setting and alerting a neurovascular specialist to suspected large vessel occlusion; images not for diagnostic use beyond notification; standalone performance AUC of 0.91. Open original source
- Supports: Peer-reviewed retrospective cohort study (222 EVT patients, July 2018–July 2022, Viz LVO implemented July 2020): median door-to-groin time fell from 109 to 75 minutes (p < 0.001); the fraction of patients receiving thrombectomy within 90 minutes rose from 26.5% to 68.1% (p < 0.001); interrupted time-series analysis confirmed the reduction (p = 0.001); no significant differences in reperfusion scores, length of stay or mortality; authors' caveats (single center, retrospective, no 90-day outcomes). Open original source
- Supports: FDA approval of the Viz LVO Stroke Platform via the De Novo pathway; creation of a new regulatory classification for computer-aided triage software; the 300-patient retrospective reader and performance study (90% accuracy, median scan-to-notification under 6 minutes, faster than standard of care in over 95% of cases, 52-minute average saving); text-notification alerting of neurovascular specialists. Open original source
- Supports: $71 million Series C (led by Scale Venture Partners and Insight Partners; total raised above $150 million); founding in 2016 by Dr. Chris Mansi, a former NHS neurosurgeon, and David Golan, a machine-learning postdoc at Stanford; San Francisco base; 2018 FDA clearance; planned expansion into cardiology, pulmonary and trauma care; Medtronic's 2019 distribution deal. Open original source
- Supports: $100 million Series D led by Tiger Global and Insight Partners at a $1.2 billion valuation; participation from Scale Ventures, Kleiner Perkins, Threshold, GV, Sozo Ventures, CRV and Susa; total venture investment above $250 million; AI detection products for aortic disease, pulmonary embolism and cerebral aneurysms; pending 510(k) for subdural hematoma; February 2022 FDA 510(k) clearance for the cerebral-aneurysm algorithm. Open original source
- Supports: European CE Mark approval (announced June 2021); the flagship product's function (AI detection of suspected large vessel occlusion on CT angiography with direct alerting of on-call stroke specialists); the large real-world multi-center study (2,544 consecutive patients, 139 hospitals, median time-to-notification 5 minutes 45 seconds, 96% sensitivity and 94% specificity). Open original source
- Viz.ai Funding Will Support Expansion of AI-Driven Viz Platform for Disease Detection and Triage Tier 2Supports: CMS New Technology Add-on Payment (NTAP) approval for the AI software; launch of the AI-driven life-sciences platform for clinical trials; first five CE Mark approvals in 2021; disease expansion (aortic disease, pulmonary embolism, cerebral aneurysms). Open original source
- Supports: The November 2024 agreement with Microsoft to launch 13 Viz.ai AI workflow models on Microsoft's Precision Imaging Network (part of Microsoft Cloud for Healthcare), covering stroke, hemorrhage, pulmonary embolism and aortic dissection, to help radiologists report more efficiently and coordinate follow-up care. Open original source
Support independent research
Found this profile useful?
You can support FactsLedger by making a donation. Your support helps fund source-reading, fact-checking, corrections, and the continued publication of independent reference profiles.
Support the research