MADI
Category: AI in Biotech / Drug Discovery
AI-CRO startup building a generative-AI digital twin platform that creates synthetic patients to replace placebo control groups in clinical trials. MADI was founded in 2025. The company is led by Kim Seong-hwan (김성환). Based in Seoul, South Korea. Team size: 11-50. Latest round: Seed. Key investors include Samsung Fire & Marine Insurance (삼성화재), Infobank (인포뱅크).
- Founded
- 2025
- Headquarters
- Seoul, South Korea
- Team size
- 11-50
Value proposition
MADI-DT generates high-fidelity AI digital twins of patients using generative AI trained on Korean/Asian clinical cohort data, reducing real patient recruitment needs, cutting clinical trial duration by ~20% (6+ months) and reducing costs by an average of ~$2.3M per trial.
Products and solutions
MADI-DT (AI Digital Twin Clinical Trial Platform, generates synthetic patients from real clinical outcome data), MADI-ACL (secondary product line referenced on company site)
Unique value
Unlike Western-data-centric competitors, MADI trains its digital twin models on Korean cohort data (40,000+ high-purity time-series clinical records: EMR, blood labs, CT/MRI), giving higher accuracy for Asian patient populations and positioning it as an AI-CRO bridge for global pharma entering Asian markets.
Target customer
Global and Asian pharmaceutical companies and CROs running clinical trials, especially those needing placebo/control arm alternatives.
Industries served
Biotech / Drug Discovery, Healthcare, Clinical Research (AI-CRO)
Technology advantage
Generative AI models trained on 40,000+ curated longitudinal Korean clinical records (EMR, Blood Lab, CT/MRI); co-founded by Chung-Ang University Hospital clinical faculty and Seoul National University AI researchers, combining clinical trial design methodology with generative modeling.
How they differentiate
Focus on Korean/Asian patient cohort data for higher regional accuracy vs. Western-data-centric competitors like Unlearn.AI; academic clinical + AI research co-founding team; MFDS regulatory approval for PROCOVA-based trial design in Korea.
Main competitors
Unlearn.AI (US, digital twin clinical trial platform, $50M Series C), other Western AI-CRO/digital twin vendors focused on non-Asian cohort data
Key partnerships
Chung-Ang University (academic/incubation partner, CES 2026 sponsor), Seoul Business Agency (SBA) CES 2026 Seoul Pavilion, Korea Ministry of Food and Drug Safety (MFDS/식약처) approved MADI's first domestic clinical trial plan applying PROCOVA methodology (2026-05).
Major milestones
MFDS (Korea FDA) approval of first domestic clinical trial plan applying PROCOVA methodology (2026-05), CES 2026 Seoul Pavilion exhibitor with Chung-Ang University support (2026-01), MOU with Chung-Ang University for medical AI talent development (2026-07), 4IR Awards (大賞) recipient.
Market positioning
Positioning as the leading AI-CRO partner for global pharmaceutical companies seeking Asian market clinical trial expansion, differentiated via localized digital twin technology.
Geographic focus
Korea and Asia-Pacific, with ambitions to partner with global pharma for US/international expansion (CES 2026 US market push).
Patents and IP
4 patent applications listed on company site (posted 2025-09): (1) Generative-AI time-series clinical digital twin / virtual control-arm patient generation system; (2) Hierarchical LLM multi-agent medical data preprocessing & missing-value imputation; (3) AI stress X-ray ACL instability quantification; (4) Bladder tumor diagnosis information apparatus/method.
About Kim Seong-hwan (김성환)
Orthopedic surgeon; Associate Professor of Orthopedics, Chung-Ang University Hospital. M.S./Ph.D. Orthopedics, Chung-Ang University; M.S. Statistics, Korea National Open University. Deputy Director, Chung-Ang University Hospital Planning & Coordination Office.
Latest news about MADI
More AI in Biotech / Drug Discovery companies
Official website: https://www.madidt.com
