Revolutionising drug discovery with AI in South-east Asia

Business Times

16-05-2025

Founder's experience
The genesis of NYB.AI was a personal one, sparked by the founder's experience with the limitations of conventional medicine and the potential of traditional remedies. This personal connection fuelled the drive to develop a more efficient and effective approach to drug discovery.
Its proprietary platform, DTIGN, launched in Q4 2024, utilises graph neural networks and molecular docking models to significantly improve drug-target interaction predictions. This technology has already outperformed second-best methods by 25 per cent, as recognised by IEEE Magazine.
SEE ALSO Sengkang General Hospital leads the way in automating sterile supply management
GET BT IN YOUR INBOX DAILY
Start and end each day with the latest news stories and analyses delivered straight to your inbox.
Sign Up VIEW ALL
DTIGN's unique approach integrates signalling pathways, reaction environments, pharmacokinetics and species-specific variations in bioactivity modelling. Unlike typical AI-driven drug discovery, which often focuses on synthetic compounds, DTIGN prioritises plant-derived compounds from South-east Asia, unlocking unique therapeutic mechanisms.
In addition to revolutionising pharmaceutical drug discovery, DTIGN's capabilities extend to other industries. The platform is being applied to the beauty sector, where plant-based bioactives are used for anti-aging, skin health and rejuvenation products.
DTIGN is also supporting innovation in animal health, including the development of solutions for zoonotic diseases. This cross-industry versatility highlights the broad impact of NYB.AI's technology and its potential to transform multiple sectors reliant on bioactive compounds.
The company's AI capabilities offer cost-effective solutions, accelerated molecule discovery and a faster time to market. NYB.AI also leverages a vast collection of billions of natural and synthetic compounds, enhancing its drug discovery process.
DTIGN's efficiency is underscored by its use of Nvidia H100 GPUs, which have dramatically reduced model training (from 80 to 10 hours) and docking (from 70 to eight hours) runtimes. AI models like AlphaFold, DiffDock and MolMIM further enhance the platform's predictive capabilities.
NYB.AI has already secured its first revenue of US$200,000 (S$261,110) from Delightex (Japan Tobacco Inc) and has established strategic partnerships with HPE, Nvidia and Equinix for AI infrastructure and go-to-market strategies. The company has also been accepted to present at Nvidia GTC 2025 and has been selected for the Nvidia Ignition Program and JLABS.
Ethical considerations and sustainability are central to NYB.AI's operations. The company operates a 6,000 square foot research lab at Nanyang Technological University, ensuring transparency, accountability and patient safety. It focuses on developing affordable treatments for complex diseases, as well as promoting healthcare equity and accessibility.
NYB.AI also supports sustainable harvesting and conservation efforts of plant-derived compounds and integrates energy-efficient AI computing.
The impact of NYB.AI's technology is significant. The company has reduced novel compound discovery timelines from 12 to 15 years to two to five years, improved bioactivity prediction accuracy by 44.83 per cent, and achieved a 64x higher hit-rate efficiency compared to traditional methods.
Projected revenue growth from US$2 million to US$10 million between 2025 and 2030 demonstrates the platform's commercial viability. Beyond quantitative impact, NYB.AI aims to enhance the sustainable quality of life by democratising access to affordable, effective treatments and supporting biodiversity conservation.
Globally iconic innovation
Its user-centric design and international recognition, including its feature in IEEE Magazine, position DTIGN as a globally iconic innovation.
NYB.AI's journey is more than a story of technological triumph – it's a testament to what is possible when empathy, science and sustainability converge.