When regulatory authorities in Europe cleared Vara's artificial intelligence system to read breast cancer scans without a radiologist looking first, the Berlin-based healthtech crossed a threshold that its competitors have approached but not reached: permission to operate as the sole first reader in a medical screening program.
The CE certification, granted under the EU's Medical Device Regulation Class IIb framework, allows the software to analyze mammograms, assign risk scores, and route patients through different review pathways entirely on its own. Vara disclosed the approval on September 2, 2024, calling it the first such clearance for autonomous AI triage in breast imaging worldwide—a claim the company has made publicly, though verification from competitors remains pending. Public filings and company announcements suggest no direct equivalent has surfaced yet.
This represents something of a philosophical shift. Most AI tools in radiology serve as a second opinion or a productivity aid, flagging suspicious areas for human experts to evaluate. Vara's newly certified platform makes the initial call itself, determining which patients need only a single radiologist review and which warrant immediate escalation to two readers—a capability validated in clinical studies including the PRAIM trial. The company said it processes more than 250,000 mammograms monthly and now supports over half of Germany's organized breast cancer screening infrastructure, according to company-reported figures from June 2024, though it declined to provide a precise current figure when asked.
CEO Jonas Muff framed the development around capacity constraints. "Speed is what European screening programs need to keep up with demand," he said in a June 2024 statement tied to a separate product approval. Germany, like much of Europe, faces radiologist shortages that have lengthened wait times and strained screening systems designed to catch cancer early.
Evidence from the Field
Vara built its case for autonomous operation on published clinical studies, including a study scheduled for January 2025 publication in Nature Medicine that found AI-supported double reading using the company's software noninferior to the standard practice of two radiologists reviewing each scan. An earlier 2023 study in the same journal showed the AI-assisted workflow maintained accuracy while cutting radiologist workload. A November 2024 analysis in The Lancet Digital Health evaluated multiple integration strategies for AI in screening programs, placing Vara's system among the approaches examined.
Still, autonomous triage introduces variables that assisted reading does not. The company said it developed ATMON, an internal monitoring system, to track the AI's performance in real time and alert administrators if accuracy begins to drift. Whether screening programs will accept that safeguard as sufficient, or demand additional human checkpoints, may determine how widely the technology spreads.
Deployments have already extended beyond Germany. Vara entered India in 2023 through a partnership with NM Medical, confirmed in the company's October 2024 funding announcement, then signed distribution agreements this past April with deepc, which operates across more than 30 countries, and Quibim, which holds exclusive rights in Spain for screening programs. The company's legal entity, MX Healthcare GmbH, showed employee counts on LinkedIn ranging from 11 to 56 as of late 2024, a bandwidth that suggests either recent hiring or uncertainty about how many people are actually on the payroll.
Funding and Founder Background
Vara closed an $8.9 million round in late October 2024, co-led by IBB Ventures' impact fund with participation from Smart Family Office in Austria and FJH in Germany. Earlier investors Merantix, EQT Foundation, and VI Partners joined. That brought disclosed total capital to roughly $20 million, following a €4.5 million raise in October 2022 and a €6.5 million Series A from OMERS Ventures in mid-2020.
Muff and CTO Stefan Bunk founded the company in 2018. Neither came from a clinical background; both arrived through Berlin's AI research ecosystem, which has produced a cluster of healthtech ventures over the past decade. Muff has described Vara's mission in terms that blend public health ambition with commercial scale, a pitch that resonates with impact-focused investors but sometimes raises questions about how the business model pencils out in markets with constrained healthcare budgets.
Ute Mercker, an investment director at IBB Ventures, said in the October funding announcement that "combining an AI-driven screening approach with proven medical evidence" aligned with the fund's criteria. Impact investors typically seek both financial return and measurable social benefit, a dual mandate that can complicate exit strategies if acquirers prioritize one over the other.
Who Else Is in the Race
ScreenPoint Medical, based in the Netherlands, holds both FDA clearance and CE marks for 2D and 3D mammography AI. The company announced $14 million in new funding plus $2 million in grants this past April, suggesting it remains well-capitalized as competition heats up. Therapixel's MammoScreen and Kheiron Medical's Mia have also secured CE marks and undergone evaluation in large-scale trials. In the U.S., where regulatory pathways diverge from Europe's, iCAD's ProFound AI maintains a foothold.
None of those firms have publicized autonomous triage certifications equivalent to what Vara received in September, though product roadmaps in this space tend to be closely held until regulatory submissions become public. The gap may narrow quickly, or it may persist if European authorities grow cautious about expanding AI's independent role.
Globally, breast cancer accounted for approximately 2.4 million new cases and 694,000 deaths in 2024, according to the International Agency for Research on Cancer. Screening programs represent a multi-billion-dollar market, but margins are thin and procurement cycles are long. Vara's ability to convert regulatory approval into sustainable revenue growth will depend on how many health systems trust the technology enough to let it make calls on its own.
For now, the company holds a certification that its competitors lack. Whether that translates into durable market advantage remains an open question.
