Healthcare / Elderly CareUS
← Sticker Wall
JOURNEY · DETAIL
AetherAI: AI Pathology Image-Assisted Diagnosis and Global Commercialization Pioneer
Founded: Ye Zhaoyuan · AetherAI Technology Co., Ltd.
JOURNEY
Key Fields
FIELD STAMPSIndustryHealthcare / Elderly Care
RegionMulti-region
ScaleMid-size
ChannelOther
Origin
Founder Ye Zhaoyuan worked in medical equipment agency in Taiwan in his early years and found that grassroots hospitals faced a shortage of over 30% in pathology doctors, with traditional pathology diagnosis suffering from low efficiency and high misdiagnosis rates. AI pathology imaging technology could significantly improve diagnostic efficiency and reduce missed diagnosis rates. In 2015, he founded AetherAI in Taiwan with an initial team of just 3 people. Starting from a single pathology slice recognition product, he targeted the rigid demand for digital and intelligent upgrades in global pathology departments, dedicating the company to building AI pathology diagnostic solutions adapted to multi-regional regulatory requirements.
Milestones
2015
Startup Stage PMF
In 2015, Ye Zhaoyuan founded AetherAI in Taiwan, assembling a 3-person technical team to develop the first AI product targeting cervical cancer pathology slice recognition. In 2017, it obtained certification from Taiwan's health and welfare department, becoming the first approved AI pathology-assisted diagnosis product in Taiwan. That same year, it completed a 30 million NTD angel round financing, serving 12 regional hospitals in Taiwan.
2019
Expansion Stage Turning Point
In 2019, AetherAI entered the mainland market, attempting to cooperate with tertiary hospitals to develop multi-disease pathology AI products. However, due to the long NMPA approval cycle in the mainland and complex hospital procurement processes, only 2 hospitals were signed in the first 18 months. Cash flow became tight, the team was reduced from 35 to 22 people, and the company was forced to contract its mainland business and pivot to cooperation with third-party testing institutions.
2021
Regulatory Breakthrough Stage Transition
In 2021, AetherAI's pulmonary nodule pathology AI product obtained China's NMPA Class III medical device certificate, becoming the third pathology AI enterprise approved domestically. That year, it established partnerships with top third-party testing institutions such as KingMed Diagnostics and DiAN Diagnostics, covering over 300 grassroots hospitals nationwide. Annual revenue exceeded 120 million RMB, a year-on-year increase of 320%.
2023
Overseas Expansion Stage Growth
In 2023, AetherAI's cervical cancer pathology AI product obtained EU CE certification, subsequently entering the Southeast Asian and Middle Eastern markets and partnering with over 10 local medical channel distributors. Overseas revenue reached 80 million RMB that year, accounting for 45% of total revenue, and the company achieved full-year profitability for the first time, with a net profit of about 12 million RMB.
2025
Globalization Layout Stage Growth
In 2025, AetherAI's products obtained US FDA certification, making it one of the few enterprises globally to hold pathology AI certifications across China, the US, Europe, and Australia. Overseas markets expanded to developed countries in Europe and the Americas, with annual revenue exceeding 580 million RMB, of which overseas revenue accounted for 68%. The post-investment valuation exceeded 3 billion RMB, and preparations for a Hong Kong IPO were initiated.
Turning Points
- In 2019, mainland market expansion encountered obstacles; the company shifted from direct sales to tertiary hospitals to cooperating with third-party testing institutions, finding a more suitable sales pathway.
- In 2021, securing the NMPA Class III certificate officially opened up the regulatory access channel for China's in-hospital market.
- In 2023, obtaining EU CE certification kicked off the global expansion strategy, leading to rapid growth in the proportion of overseas revenue.
- In 2025, capturing US FDA certification established the company as one of the pathology AI enterprises with the most complete global regulatory footprint.
Failures & Pitfalls
- In 2019, when attempting to directly interface with mainland tertiary hospitals, due to long approval cycles and slow payment collections, the company had no new major clients for 18 consecutive months, burning through 20 million in financing.
- In 2020, the multi-disease general pathology AI model developed by the company suffered a 12% missed diagnosis rate in clinical tests due to data bias issues, forcing a complete overhaul and redesign, which delayed the launch plan by 6 months.
- In 2022, upon entering the Japanese market, because it failed to adapt in advance to the staining standards of Japanese pathology slices, the initial batch of 100 sets of products was entirely returned, resulting in losses of over 3 million.
关键成功要素
- Secure regional market regulatory certification first before gradual expansion to avoid cross-regional compliance risks.
- Focus deeply and thoroughly on core pathology scenarios, avoiding the blind expansion of disease types that disperses resources.
- Lightweight product design reduces hospital deployment costs and enhances grassroots market penetration.
- Localized service teams paired with local channel distributors rapidly adapt to the medical system rules of different countries.
Lessons
- Regulatory access for medical AI products is a prerequisite for commercialization; certification qualifications for target markets must be planned in advance.
- The medical industry has a long decision-making chain; directly engaging terminal hospitals requires sufficient cash flow reserves, and prioritizing channel partners can reduce payment collection risks.
- There are large differences in pathology sample standards across regions; product localization adaptation must be completed before going overseas to avoid compliance and quality pitfalls.
- Focusing deeply on a single disease type is easier to get through regulatory approval than generalized products and is also more easily accepted by clients.
Core Data
- revenue_2025：580 million RMB
- overseas_revenue_ratio_2025：68%
- npmpa_certificates：3
- overseas_certificates：12
- valuation_2025：3.0 billion RMB
Competitors / Peers
AetherAI's main competitors in the domestic pathology AI track include DiAN Diagnostics, the AI pathology department of KingMed Diagnostics, and cross-disease AI medical enterprises such as Deepwise and Infervision. In overseas markets, it faces competition from leading US pathology AI vendors such as Paige.AI and PathAI. Compared to competitors, AetherAI's advantage lies in simultaneously holding multi-region regulatory certifications across China, the US, Europe, and Australia, with product deployment costs at only 1/3 of overseas competing products, giving it a prominent cost-performance ratio in the grassroots market.