Singapore firm raises $50m to advance AI-driven clinical trial automation

Deep Intelligent Pharma (DIP), a Singapore-based leader in medical AI, has secured $50 million in Series D funding. The new capital is earmarked for the iterative development of its autonomous clinical systems and the expansion of its global commercialization network, aiming to shift AI’s role from a simple assistant to essential decision-making infrastructure in drug R&D.

“Atomic Agent” architecture and industrial-grade precision The core of DIP’s innovation lies in its Synaptic Agent Ecosystem, which mimics human brain bionics:

  • Logic decomposition: The system breaks down complex clinical workflows into 10,000 high-precision “atomic agents.” Each agent specializes in specific tasks, such as verifying inclusion/exclusion criteria or simulating statistical endpoints.

  • Eliminating hallucinations: Trained on “dark data” from tens of thousands of past projects, the system achieves a 99.9% accuracy rate. This eliminates the “hallucination” problems typically found in generic AI models when applied to highly regulated medical environments.

Real-world impact: The Japan case study DIP’s “simulate first, execute later” strategy was recently validated through its collaboration with Immunorock, a Japanese cancer vaccine developer. Using DIP’s autonomous platform for a Phase I/IIa immunotherapy trial, the project achieved:

  • Zero-revision approval: The clinical protocol passed the Japan PMDA review in a single submission with no revisions required.

  • Resource optimization: The process saved 90% of the usual development time and allowed the team to proceed without relying on traditional Contract Research Organizations (CROs).

The Series D round was led by CDH Baifu, with continued support from Xinding Capital and HongShan Capital.

Source: https://www.bioxconomy.com/clinical-and-research/deep-intelligent-pharma-raises-50m-to-advance-ai-clinical-trial-automation-technology

0 0 votes
Article Rating
Subscribe
Notify of
0 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments