Biomy
Biomy makes DeepPathFinder, cloud software that uses AI to map immune and cancer cells on tissue slides for drug research. Drug companies, contract research organizations and research groups use it to look for patterns that could become biomarkers: signs that predict which patients will respond to a treatment.
The software finds and counts cells on standard and multi marker stained slides, and studies how they cluster within a tumor. It covers cancer immunotherapy research and autoimmune and inflammatory disease. It is sold for research only, not for clinical use. Biomy says drug companies and research institutions use it, without naming them.
Since January 2025, Biomy and NEC have marketed it together as a platform for biomarker discovery. Biomy also works with Chugai Pharmaceutical on an AI program to support cancer diagnosis, and the contract research organization Shin Nippon Biomedical Laboratories formed a capital alliance with it in 2024.
Biomy is a Tokyo company founded in 2019. No price is published.
Capability Axes
An AI Health Index grade measures what a buyer can verify from public sources on the date shown. It is not a rating of how good the product is. A vendor can build an excellent system and grade low on an axis because it publishes nothing an outsider can check. How grades read
The models are the product. DeepPathFinder is deep learning analysis of pathology slides, finding and counting immune and cancer cells and their spatial patterns. The biomarker discovery services Biomy offers with NEC are built on it. Biomy sells no scanner, laboratory service or data product that would stand without the models.
Ask which model version produced a given analysis, and whether models are retrained for a customer's stains and scanners.
Biomy publishes no oversight structure. DeepPathFinder is labeled for research use only, which keeps it out of patient care. But Biomy describes no review of its cell detections or biomarker outputs, and no confidence threshold. Nothing says how a researcher should check a result before relying on it in a study or in deciding how to group patients.
Ask how detections are reviewed, what confidence measures come with outputs, and how errors are reported and fixed.
Biomy describes its approach in general terms, without identifying anything. It calls DeepPathFinder a deep learning platform for cell detection and spatial analysis. It also describes a government funded project to combine pathology images with spatial gene activity data using generative AI.
No model architecture, training data, version or update practice is published.
Ask for the model architecture, training data sources and size, the version in use, and how updates are announced.
Biomy develops DeepPathFinder's models itself with academic collaborators, and no outside model provider is named or implied. It names AWS, Microsoft and NVIDIA as technology partners, without saying which hosts customer data.
When it is sold with NEC, NEC's image analysis is combined with Biomy's, without saying who processes which images. No subprocessor list is published.
Ask which parties process uploaded slides, including the cloud host and NEC, and under what terms.
Biomy reports use without names or figures. It says drug companies and research institutions use DeepPathFinder. It names academic collaborators at the University of Tokyo and at Keio, Akita and Juntendo universities, and refers to conference abstracts at ASCO 2023.
No customer, validation cohort, agreement with pathologists or performance figure is set out on its site or in its announcements.
Ask for cell detection accuracy against pathologist annotation, the studies behind the ASCO 2023 abstracts, and a reference customer.
Nothing Biomy publishes covers how patient data moves through the product. Its privacy policy promises appropriate security for personal information collected through its website.
Nothing says how long uploaded slides are kept on the cloud platform, whether they train Biomy's models, or how training data from its academic partners was de identified. No safety engineering for the models is published.
Ask how long uploaded slides are kept, whether customer images train any model, and how partner training data was de identified.
Biomy states no HIPAA position, and no privacy document covers the product. Its privacy policy, established in November 2019 under Japan's Act on the Protection of Personal Information, covers website visitors and inquiries. It does not address pathology images or patient data processed in DeepPathFinder.
No HIPAA position, business associate agreement or data processing agreement is published.
Ask how patient data in uploaded slides is handled, whether images must be de identified first, and for the data processing terms, including for US customers.
Biomy publishes no security audit or certification. Its privacy policy promises appropriate security measures. No ISO/IEC 27001 certificate, SOC 2 report, penetration testing statement or trust center appears on its site. Certifications held by its cloud partners would be the host's, not Biomy's.
Ask for any security attestation covering DeepPathFinder, with its scope and date.
Biomy makes no device claim, and the product is labeled for research. Biomy states that DeepPathFinder is for research use only and not for clinical use. The FDA's databases list no Biomy device.
Its memorandum with Chugai concerns developing a program to support cancer pathology diagnosis. Its partnership with NEC, which holds a Japanese license to make and sell Class I medical devices, includes support for device approval. But no Biomy product is presented as cleared or approved anywhere.
Ask whether any Biomy product is planned for submission to Japan's PMDA or the FDA, and under which partner's license.
Biomy refers to research, but publishes no evaluation across groups and no governance process. It points to conference abstracts and academic collaborations. It publishes no results by tissue type, stain, scanner, laboratory or patient group, and describes no process for evaluating models before release or monitoring them afterward.
Ask for performance by tumor type, stain and scanner, and how model updates are evaluated.
Biomy publishes nothing on what happens when the system is wrong. No error rate, review step, warranty, remedy or correction route for DeepPathFinder outputs appears on its site, and the product is labeled for research use only.
Ask what a license commits Biomy to when an analysis is found wrong, and for error rates on the cell types you would count.
Biomy publishes no integration. DeepPathFinder is a cloud platform for research analysis. No image management system, scanner, laboratory system, data format or interface is named for bringing slides in or sending results to a customer's research systems.
Ask which slide formats and image management systems DeepPathFinder reads, and how results are exported.
DeepPathFinder is described as cloud based, and where it runs is implied, not stated. Biomy names AWS, Microsoft and NVIDIA among its technology partners. Nothing published says which cloud hosts customer slides, in which region, or whether images leave Japan.
Ask for the hosting provider and region, and whether data from customers outside Japan is processed in Japan.
No price is published, but the outline of how the product is sold can be worked out. DeepPathFinder is offered to drug companies and laboratories as cloud software and through biomarker discovery services. Since January 2025 it has been marketed together with NEC.
No license basis, per slide or per study price is published.
Ask whether DeepPathFinder is licensed per user, per slide or per study, and how pricing differs when bought through NEC.
Biomy states where DeepPathFinder applies, with nothing on its pages validating it. It analyzes H&E and multiplex immunohistochemistry slides for tumor microenvironment research, in cancer immunotherapy and in autoimmune and inflammatory disease, for research use only.
Biomy sells biomarker discovery for drug development, alone and with NEC, which is why it is also listed under drug discovery. No validation by tissue type, stain or disease is published.
Ask which tumor types, stains and scanners DeepPathFinder has been validated on.
Pricing
Vendor-published figures are labeled as such. Figures labeled “Estimated” are derived from third-party sources and have not been confirmed by the vendor.
| Entry Price | Pricing Basis | BAA Tier | Implementation | Source |
|---|---|---|---|---|
|
Not published
|
Cloud research software (DeepPathFinder) and biomarker discovery services for biopharma, sold directly and with NEC; no unit of charge or price published. | Not published. The privacy policy covers website and inquiry data; no data processing or business associate agreement is published. | Not published. | Vendor Published |
Biomy describes DeepPathFinder as research use software for biopharma and laboratories and offers biomarker discovery services, including jointly with NEC since January 2025, without a price, unit of charge or license term. A buyer should ask for the license basis and how pricing differs through NEC.