Nature’s Solutions for All
The healing power of nature has been recorded for millennia, in books spanning hundreds of thousands of pages, from different cultures. They are continually studied in the modern ages as published in medical and scientific journals. Our AI engine, Ligand-Protein Repurposing Discovery and Search Engine (LPROSE) has been trained with those written documents and can accurately suggest medicinal plant extracts for any health problems.
Reinventing the herbal medicine research and development
Accurate & Safe
Accuracy and precision in our design process translate into high product safety. Rest assured that our AI system only includes highly effective ingredients in our formulations and no unnecessary compounds are added. Our “Formulated Using AI” tagline guarantees our products are free from any steroids and controlled substances, unlike many herbal remedies in the market.
Fast Development
The AI engine can help to design better and more precise medicine faster than using conventional wisdom. Since we rely on naturally available chemical compounds, we are also faster than other synthetic drug designers. Less trial-and-error process means we are better at responding to sudden spike in demands or trends.
Cost Effective
High-quality products cost a lot of money. Our AI system can help to reduce the R&D process which usually relies upon much technical expertise. The savings in R&D go straight into reducing our health products’ price.
Thinking of launching new herbal medicine, supplement, cosmetic or functional food?
We Can Help You Find the Right Formulation!
Our Technology
Dataset
Our AI engine has been trained on an integrated dataset of almost 400,000 natural products from around the world, notably from traditional Chinese medicine, Indian ayurvedic, African herbal medicine and many others. The whole dataset contains encyclopaedic information on more than 10,000 plants and fungi. The dataset is complemented with data from 3000 medicinal targets of human protein and pathogenic microbes.
Understanding
The dataset is analysed by natural language processing which is based on Bidirectional Encoder Representation from Transformers (BERT) created by Google.
Recommendation
A state-of-the-art Recommendation System is performed using output from GNN. This will produce different formulations with probability scoring showing its potential effectiveness.
Insight
The Graph Neural Network (GNN) is built to map the relations between plants, phytochemicals, peptides, human diseases, protein targets and various other parameters.
Herbal Remedy Formulation
A human-in-the-loop is introduced at the final stage where we will engage with various partners to test our most promising formulations.
Our Technology
Dataset
Our AI has been trained on an integrated dataset of almost 400,000 natural products from around the world, notably from the traditional Chinese medicine, the Indian ayurvedic, the African herbal medicine and many others. The whole dataset contains encyclopaedic information of more than 10,000 plants and fungi. The dataset is complemented with data of 3000 medicinal targets of human protein and pathogenic microbes
Understanding
The dataset is analysed by natural language processing which is based on Bidirectional Encoder Representation from Transformers (BERT) created by Google.
Insight
The Graph Neural Network (GNN) is built to map the relations between plants, phytochemicals, peptides, human disease, protein target and various other parameters.
Recommendation
A state-of-the-art Recommendation System is performed using output from GNN. This will produce different formulations with probability scoring showing its potential effectiveness.
Herbal Remedy Formulation
A human-in-the-loop is introduced at the final stage where we will engage with various partners to test our most promising formulations.