Biofourmis

Positioning: care-at-home with a clinical analytics engine — and a turbulent corporate history

Biofourmis sells into one of the hardest problems in healthcare delivery: keeping high-acuity patients out of the hospital without lowering the standard of care. The pitch is a combination of medical-grade wearable biosensors, a machine-learning analytics engine called Biovitals that builds a personalised physiological baseline for each patient, and the clinical service layer — virtual nursing, care navigators, in-home services — that acts on what the algorithms surface. Health systems buy it to run hospital-at-home admissions, post-discharge monitoring and chronic disease programmes; until recently biopharma bought a separate line for decentralised clinical trials. It is enterprise software wrapped around clinical labour, and evaluating it means evaluating both.

Readers should have the corporate timeline before the product details, because it is unusually eventful. Founded in 2015 (originally Singapore-based, with Kuldeep Singh Rajput and physician Maulik Majmudar among the founders), the company relocated to Boston in 2019, raised roughly $460 million across six rounds — including a $300 million Series D in 2022 at a $1.3 billion valuation, backed by SoftBank Vision Fund, General Atlantic, Bessemer, Intel Capital and CVS Health — and peaked around 650 staff. Then: 120 roles cut in July 2023 (about 15%), the founding CEO stepping down a month later, an all-stock merger with CopilotIQ in October 2024 that put CopilotIQ's chief executive David Koretz in charge with close to $100 million of fresh investor money, and the January 2025 sale of the life-sciences division (Biofourmis Connect) to ActiGraph. The Biofourmis name and product lines continue under the combined entity, but a buyer in 2026 is contracting with a reshaped company.

What it actually does

The core of the platform is Biovitals, the analytics engine that ingests continuous data from clinical-grade wearables plus patient-reported outcomes and health history. Its distinguishing technical claim — repeated across the product material and worth verifying in any procurement — is that it models each patient against their own physiological baseline rather than a population norm, and raises an alert when that individual drifts from their own pattern. Vendor material says it analyses 120+ biomarkers in real time and can flag deterioration days before a critical event; the "days before" figure is a vendor claim and I would ask for the study behind it. BiovitalsHF, the heart-failure application of the engine, holds FDA Breakthrough Device Designation, and the analytics carry FDA clearance — the strongest regulatory signals in the RPM category, and the main reason Biofourmis shortlists above smaller competitors.

Around the engine sit the care-delivery products. Biofourmis Care is the configurable care-at-home platform for hospitals, health systems and payers, spanning acute (Hospital@Home — inpatient-level care at home with continuous monitoring, virtual nursing and on-demand in-home visits), post-acute (post-discharge and post-surgical monitoring) and chronic care (including oncology monitoring via the Gaido Health lineage). It is device-agnostic on the sensor side and integrates with Epic EHR; a February 2024 GE HealthCare partnership was designed to extend that manufacturer's patient monitoring from the hospital into the home. The service model matters commercially: this is not pure software. Licensed navigators and virtual specialty care are part of the delivery, and in-home services (diagnostics, DME, medications, transportation, tech support) are orchestrated through the same clinician dashboard. The CopilotIQ merger added high-frequency connected care for older adults with hypertension and diabetes — continuous biomarker data plus behavioural analytics plus nurse visits — which is what extends the combined platform down the acuity spectrum to pre-surgical optimisation and everyday chronic management.

My testing notes

First concrete observation: the product boundary moved twice in fifteen months, and the seams are visible. If you were evaluating Biofourmis for decentralised clinical trials, that business is now ActiGraph's — the Connect division, with its 30-odd proprietary AI-driven algorithms and digital measures work (including a nocturnal-scratching measure for atopic dermatitis), transferred in the January 2025 transaction and Biofourmis no longer sells it. Care delivery is what remains. I checked the newsroom directly: the AdventHealth acute care-at-home selection (January 2025) and the Lee Health hospital-at-home expansion (November 2024) are the reference deployments that anchor the current story, alongside named relationships with UCI Health, Community Health Network, Augusta University Health, Orlando Health and WellSpan. More than 50 health systems and payers is the company's own customer-count claim.

Second observation: the personalised-baseline approach is the part of the product I would test hardest in evaluation, because it determines alert volume, and alert volume determines whether nurses keep using the system. Population-threshold RPM generates noise; a per-patient drift model should cut it. But the model is only as good as the wearables feeding it and the escalation pathway behind it — and Biofourmis explicitly sells that pathway (clinical teams, navigators, in-home services) as part of the package. When I map this against other RPM deployments I have looked at, the failure mode is almost never the algorithm. It is what happens at 2am when an alert fires and nobody owns the response. During procurement, I would scope the clinical staffing agreement with the same care as the software licence.

Third observation, on stability signals: the funding and restructuring history is a mixed record. The $1.3 billion valuation and blue-chip investor list are real, and the CopilotIQ merger injected close to $100 million and a clear strategic refocus on in-home care delivery. But the 2023 layoffs, the founder-CEO departure, the all-stock structure of the merger and the divestiture of the life-sciences arm all landed within 18 months. Health-system contracts run multiple years; I would want explicit continuity clauses and a named executive sponsor going in.

What it does not do

There is no self-serve entry point and no public pricing of any kind. Contracts are custom-negotiated with health systems, IDNs and payers, priced around patient volume, programme type (Hospital@Home, chronic, post-acute), integration scope and term — and typically deployed under value-based arrangements where cost ties to outcomes and readmission reduction. That model suits large buyers with real patient volume; a small practice cannot buy this product.

It is also no longer a clinical-trials vendor. Biopharma buyers who knew Biofourmis for digital endpoints must now work with ActiGraph. And some claims deserve a "prove it" flag: the 120+ biomarker count, the days-of-early-warning figure and the readmission-reduction outcomes are vendor-reported, and independent head-to-head comparisons against Current Health, Cadence or Vivify Health are scarce. Third-party review aggregators score it in the mid-7s out of 10, but those panels synthesise marketing material rather than running clinical pilots, so I weight them lightly. Finally, the AI layer's posture is detection and decision support — it surfaces risk and drives workflows. Clinical judgement on medication changes and admissions stays with the humans, as it should, and the digital therapeutics pipeline (software-based therapy for heart failure) has been quiet relative to its early billing.

Verdict

As a clinical monitoring platform, Biofourmis is one of the most serious products in the category: FDA-cleared analytics with a Breakthrough-designated heart-failure application, a personalised baseline model that is technically differentiated, Epic integration, and a service layer that answers the "who responds to the alert" question that kills most RPM programmes. The named health-system deployments — AdventHealth, Lee Health and others — are the right kind of evidence. Against that, you are buying from a company mid-reinvention: a merger that changed the CEO, a divestiture that halved the product story, layoffs in the recent past, and no public commercial terms to anchor a negotiation. I rate it 3.6 — strong on core clinical functionality and improving on ecosystem, held back by corporate continuity risk and buyer-invisible pricing. If you are a health system standing up hospital-at-home at scale and you can negotiate continuity protections, it belongs on the shortlist. If you want transparent software pricing or a clinical-trials platform, look elsewhere.

Sources

• https://biofourmis.com/news-insights
• https://insurtechlist.com/companies/biofourmis
• https://aihealthindex.io/vendors/biofourmis
• https://startupintros.com/orgs/biofourmis
• https://breldigital.com/companies/biofourmis
• https://aicentralresources.com/tool/biofourmis
• https://aicentralresources.com/compare/biofourmis-vs-vizai
• https://crosstreecapital.com/hubfs/theme-crosstree/pdf/Crosstree%20Deal%20Insights%20One-Pager%20ActiGraph%20+%20Biofourmis.pdf?hsLang=en
• https://healthaicentral.com/companies/biofourmis
• https://topreviewed.ai/products/biofourmis
• https://galaxy.32dots.de/company/biofourmis-697f9ac049b197ec32d8cdee