AI Platform 

Metrics-driven, approach agnostic algorithms 

Proprietary domain-specific models and scoring mechanisms, with algorithms combining deep learning techniques and molecular docking simulations for cross-verification.

Cost-effective, purpose specific hardware infrastructure

Access to high performance/cost-efficient infrastructure through AI Bridging Cloud Infrastructure and in-house NVIDIA GPU servers for prototyping and testing. 

Data curation and analytics

We are using a combination of open-source large databases, such as ChEMBL, PubChem, literature searches, to initialize model training. BrightCore is creating a curated database and analytics layer, through synthetic data generation, building proprietary features and analytics as elements for algorithms on top of the data layer.