For pathologists, the ritual is familiar: slice tissue, stain it with chemicals, wait. Sometimes days. Sometimes weeks. The results determine cancer treatment plans, guide clinical trials, and consume tissue samples that can never be replaced.
ViewsML Technologies thinks it has a shortcut.
The Vancouver startup closed a $4.9 million seed round on April 20, 2026, led by Wittington Ventures and backed—notably—by Mayo Clinic and Continuum Health Ventures. The funding, which the company says was oversubscribed, will bankroll commercial expansion and clinical validation work for ViewsML's central pitch: machine learning that predicts biomarker expression from standard pathology slides, no staining required.
It's a seductive premise in a field where tissue is scarce and timelines are measured in patient anxiety. Whether it holds up under clinical scrutiny is the question investors are now betting on.
Virtual Stains, Real Ambitions
ViewsML's platform starts with the most routine artifact in pathology—hematoxylin and eosin slides, the pink-and-purple stained tissue slices that pathologists have relied on for over a century. Using what the company describes as "virtual immunohistochemistry," its AI generates predictions for where specific biomarkers would appear at the cellular level, skipping the physical staining process that normally requires precious tissue and specialist labor.
The result, in theory: spatial, quantitative biomarker data in minutes rather than the days or weeks traditional immunohistochemistry demands.
Founded in 2022 by Kenneth To—who previously co-founded Wax-it Histology Services and worked at Eli Lilly and STEMCELL Technologies—the 12-person company frames its technology as a virtual biomarker library. Its outputs, according to the company's website, slot into existing digital pathology systems. The target markets are translational research, contract research organizations, and eventually diagnostics, though ViewsML hasn't disclosed which customers are paying or how much revenue it's generating.
Why Mayo Matters
Mayo Clinic's participation carries weight. The Rochester-based health system doesn't casually write checks to early-stage AI companies, and its presence in this round signals institutional validation that venture investors crave. The press release confirmed Mayo Clinic's involvement but did not specify the investing entity—the health system has multiple arms—but industry publication HIT Consultant called it "significant strategic participation."
Mayo launched its own digital pathology initiative last year, part of a broader modernization push across anatomic pathology. The implication: one of the country's most respected medical institutions sees something here worth exploring, even if it's not yet endorsing the technology for clinical use.
Wittington Ventures partner Zeeshan Ali expressed in the announcement that ViewsML's approach "has the potential to transform how biomarker testing is conducted." Neither firm disclosed valuation, a common omission at this stage but one that leaves outside observers guessing whether the round priced ViewsML as a moonshot or a measured bet.
A Year of Validation Theater
ViewsML's recent history reads like a startup checking boxes—partnerships announced with careful cadence, conference appearances timed for visibility. Providence Health Care Ventures signed on in September 2025. Contract research outfit iProcess Global Research followed in August. Swiss biopharma Debiopharm came aboard in June for translational research work. An earlier collaboration with Dartmouth Hitchcock Medical Center was established in April 2025.
The company showed up at the United States and Canadian Academy of Pathology annual meeting in March, then again at the American Association for Cancer Research gathering in April, presenting posters that referenced pilot validation work with HistoWiz, among others.
It's the kind of activity calendar that suggests a company building credibility one study at a time. Whether any of these partnerships translate to recurring revenue or just more pilot data remains undisclosed.
The Broader Bet on Virtual Staining
ViewsML isn't alone in chasing this opportunity. Pictor Labs recently announced on-premises deployment options for its virtual staining suite. OptraSCAN offers competing virtual IHC technology. Academic literature continues churning out feasibility studies on AI-generated stains, though rigorous, head-to-head comparisons in peer-reviewed journals are harder to find.
The appeal is straightforward: tissue samples from biopsies and surgeries are finite, often irreplaceable. Every physical stain consumes material that could be used for another test. If an algorithm can predict what those stains would show, it preserves tissue for additional analysis and collapses turnaround times that research labs and hospitals complain about constantly.
The catch, as with most AI-in-medicine pitches, is proving clinical utility at scale. Algorithms trained on one population or tissue type don't always generalize. Regulatory pathways for diagnostic applications remain hazy. And pathologists—understandably—tend to trust physical stains they can see under a microscope more than computational predictions, however accurate the validation studies claim to be.
What Comes Next
ViewsML says the new capital will fund expansion of its virtual biomarker library, deeper clinical validation, and hiring across engineering, science, and commercial functions. The company describes partnerships with "leading" pharmaceutical companies, hospitals, and diagnostic labs, though it hasn't put numbers to those claims.
For now, the startup is wagering that pressure on biospecimen resources and the glacial pace of traditional workflows will eventually tilt the market in its favor. Mayo Clinic's involvement suggests at least one major institution is willing to explore that thesis.
Whether virtual stains become standard practice or remain a niche research tool will depend on data that hasn't been published yet and regulatory clarity that hasn't arrived. The $4.9 million buys ViewsML time to generate both.
