Six 2025 Health-Tech Startups Facing the Proof Test

Function Health, Nabla, Ambience Healthcare, Hippocratic AI, Berry Street and Fay reached meaningful commercial milestones in 2025. That momentum keeps them relevant, but their harder test is whether expanding products and services produce measurable clinical, operational or access benefits.

The six companies represent consumer testing, clinical documentation, patient-facing AI and insurance-supported nutrition care. Their financing and deployment milestones show demand, not proof of better outcomes, so this is a watchlist rather than a product endorsement, clinical recommendation or investment ranking.

Why these six companies still matter

The defining health-tech development behind this selection was the movement of software into ordinary healthcare workflows. These companies are interpreting laboratory results, drafting clinical notes, preparing patients for appointments and helping registered dietitians operate insurance-funded practices.

That shift raises the standard by which they should be judged. Capital can finance expansion and customer growth can demonstrate demand, but neither establishes diagnostic value, improved health outcomes or sustainable savings. Each company below now faces a visible real-world test that goes beyond attracting users or investors.

1. Function Health must turn more testing into useful decisions

Function Health sells direct access to a broad set of laboratory tests rather than requiring every test to originate in a conventional primary-care visit. Its current membership page lists a $365 annual price, more than 160 tests per year, twice-yearly and on-demand testing, clinician review, optional MRI and CT scans, and beta access to Private AI Chat and personalized Protocols.

The attraction is longitudinal access: members can keep results together and compare measurements over time. The harder question is whether a broad panel produces useful decisions rather than incidental abnormalities, unnecessary follow-up or false reassurance. Evidence on how often flagged results change care—and how easily outside clinicians can use the records—would be more informative than the number of measurements offered.

2. Nabla is moving beyond ambient documentation

Nabla began with an immediately legible use case: listening during a consultation and helping prepare the clinical note. A June 2025 financing update details a $70 million Series C, $120 million in total funding, deployment across more than 130 healthcare organizations, use by 85,000 clinicians and planned expansion into coding, clinical-documentation integrity and commands that initiate actions in electronic health records.

That wider scope changes the risk profile. Drafting text for clinician review is different from initiating an EHR action or influencing a billing code. Correction rates, performance across specialties and languages, and the amount of review still required will therefore matter as much as headline time savings.

3. Ambience Healthcare is testing whether one platform can span a hospital

Ambience Healthcare is pursuing documentation, coding and workflow support across outpatient clinics, emergency departments and inpatient care. Its July 2025 Series C page records a $243 million round and describes a platform operating across more than 100 specialties with integrations for major electronic-record systems.

The consequential question is whether that breadth survives local complexity. Emergency medicine, an inpatient ward and a scheduled specialist visit have different vocabulary, pace, coding requirements and tolerance for error. Independently verifiable results on documentation quality, clinician workload, denied claims and maintenance costs would reveal more than the nominal number of supported specialties.

4. Hippocratic AI puts generative agents in front of patients

Hippocratic AI presents the most direct safety challenge in this group because its agents communicate with patients rather than operating solely behind the scenes. TechCrunch’s January 2025 coverage documents a $141 million Series B at a $1.64 billion valuation, non-diagnostic uses such as preoperative preparation and remote monitoring, and contracts with 23 health systems and insurers during 2024.

Keeping the product non-diagnostic is an important boundary, but it does not remove risk. Patients may misunderstand instructions, fail to escalate worsening symptoms or disclose sensitive information during an automated exchange. Completion and escalation rates, performance across languages and health-literacy levels, and explicit rules for transferring a conversation to a qualified person are central measures of safety.

5. Berry Street is building infrastructure for independent dietitians

Berry Street approaches health technology through provider capacity rather than diagnosis or clinical AI. TechCrunch’s February 2025 account places Berry Street and Fay at $50 million each for their respective rounds, describes both as platforms connecting patients with independent registered dietitians, and records Fay’s $500 million valuation and the role of GLP-1 treatment in increasing demand for nutrition services.

Berry Street gives dietitians tools for running independent practices, including arrangements that can add appointment capacity outside clinicians’ primary jobs. Its opportunity depends on making insurance verification, scheduling, documentation and reimbursement workable for providers while preserving continuity for patients. Completed appointments, time to first visit, provider retention and claims outcomes would provide a clearer view of access than practitioner sign-up totals.

6. Fay is scaling an insurance-backed nutrition marketplace

Fay also matches patients with independent registered dietitians, placing it in direct competition with Berry Street and other virtual nutrition platforms. Its rapid financing sequence makes execution more informative than another funding milestone: the service must consistently turn insurance eligibility into completed, affordable care.

The model could widen access to medical nutrition therapy, particularly for patients who do not know that their plan may include it. Coverage still varies by insurer, diagnosis, clinician and plan design, so “covered” should not be treated as synonymous with universally free. Fay’s durable value will depend on accurate benefit checks, understandable billing, sustainable economics for dietitians and engagement beyond an initial appointment.

The evidence that would separate momentum from impact

Although the companies address different problems, four questions separate a promising deployment from dependable healthcare infrastructure:

Does the product improve a defined outcome? Time saved, completed follow-ups, fewer billing errors and appropriate escalation are more useful than registration totals alone.
Who checks the technology? Patient-facing guidance and actions inside an electronic record require explicit human oversight, audit trails and clear limits on automation.
Does access survive the payment process? A service is not meaningfully accessible if eligibility is unclear, claims frequently fail or clinicians cannot sustain participation.
Can performance claims be independently examined? Company materials can establish product scope and deployment scale, but health-system evaluations and peer-reviewed studies are needed to demonstrate broader impact.

These standards explain why the six companies remain worth watching after their 2025 milestones. They have moved far enough into real healthcare settings to be evaluated on consequences; the decisive evidence will come from what happens to patients, clinicians and costs after deployment.

Six Health-Tech Startups From 2025 Now Face the Hard Part: Proving Impact