Human Array Platform (HAP) 

Traditional drug discovery and toxicity studies often rely on animal models that fail to fully capture human biology. Our platform overcomes this gap with a suite of in-house developed real-time biosensors integrated human derived organoids and 3D cell cultures

Time-series forecasting 

Simulate future cycles of drug effect on organoids (e.g. dose–response curves evolving over time)
Predict rebound, adaptation, toxicity onset, recovery curves 

Stratified modeling (population effects) 

Model how ethnicity, gender, age, genetic variants influence responses 
Virtual cohorts of in silico organoids to explore variability 

Mechanistic inference & feature attribution 

Interpret which pathways / readouts drive observed effects 
Suggest mechanistic biomarkers or secondary endpoints 

Active learning & experiment design 

The AI suggests next experiments to run to maximally reduce uncertainty 
Optimization of dose, schedule, combinations 

Predictive translation to in vivo & clinical 

Map simulated organoid response curves to predicted human PK/PD, toxicity, therapeutic windows 
Project human outcomes, risk scores, stratification for trial design 

Data Pipeline 

Edge acquisition with device identity and time synchronisation; encrypted storage; audit trails; rolebased access. 

Feature Engineering & Models 

Trajectory based features (rates, inflections, coupling of pH with analytes) feed supervised models for toxicity classification and stress phenotyping. Model cards document training data, validation, and intended use. 

Interoperability 

APIs for ELN/LIMS; export in standard formats; FAIRaligned metadata. 

Compliance Readiness 

Designed for 21 CFR Part 11 readiness and alignment with emerging NAM guidance.