hema.to

"AI diagnostic support for haematology labs, making decisions more than twice as fast while maintaining quality." [1]

Where it wins

• Speed vs. manual gating: hema.to’s AI engine cuts time to report by automating the gating and interpretation workflow, enabling labs to increase throughput significantly. [1]
• Regulatory certainty: Unlike many AI tools that are research-use-only, hema.to’s B-NHL engine holds CE-IVD (IVDD) certification, allowing for clinical deployment in diagnostic settings. [1]
• Transparency in AI: The system explains the reasoning behind its decision support, allowing cytometrists to configure, adapt, and export medical reports in seconds rather than typing them manually. [1]

Credibility: Performance studies show high concordance with 6,500+ routine results across key populations including granulocytes, lymphocytes, B cells, NK cells, and T-cell subsets (CD4+, CD8+). [1]

• AI-driven diagnostic workflow: The company sells an AI engine that integrates into existing cytometry workflows to automate gating and interpretation. [1]
• Dual-track product strategy: Offers a CE-IVD certified product for clinical use and a RUO product for research, allowing for phased adoption by labs. [1]
• High-margin software delivery: The value is delivered via software, scaling with lab throughput without the need for physical hardware or consumables. [1]

• Traditional cytometry software: Competitors offer manual gating tools that lack AI-driven automation, resulting in slower workflows. [1]
• Emerging AI diagnostics: Other AI tools may offer speed but lack the CE-IVD certification required for clinical use in the EU. [1]
• Differentiators: hema.to’s combination of CE-IVD certification, high concordance with expert results, and transparent reasoning sets it apart in the clinical haematology space. [1]

• Slow manual gating: Cytometrists spend significant time on the manual process of gating and interpreting cell populations, slowing down lab throughput. [1]
• Tedious report typing: Manual creation of medical reports is time-consuming and prone to error, taking away from clinical decision-making. [1]
• Inconsistent results: Variability in human interpretation can affect the quality and consistency of diagnostic results across different operators. [1]

The CE-IVD certification is a significant moat, allowing hema.to to deploy in clinical settings where many AI competitors cannot. The focus on haematology is a strong wedge, but expansion into other areas of pathology could unlock larger markets. The main risk is regulatory scrutiny as AI in healthcare becomes more regulated; maintaining compliance will be critical. The next signal to watch is the adoption rate of the RUO product as a pipeline for future clinical deployments.

Expand the RUO product line to include more research applications, creating a larger funnel for clinical adoption. Develop a self-service onboarding process for smaller labs to reduce the cost of customer acquisition and support. Publish more case studies from diverse lab settings to demonstrate the AI’s robustness across different patient populations. Explore partnerships with cytometry hardware manufacturers to integrate hema.to directly into their software ecosystems.