Not just screenshots. Experiments you can test.
Interactive engineering demos with server-side models, public-data experiments, transparent metrics and privacy-safe business analytics.
When the API is online, inference comes from the server. The browser does not need to train any model locally.
Historical Real-Estate Price Intelligence
Historical asking-price model, prediction interval, model metrics and feature importance.
Persian Language Intelligence
Sentiment, intent, keywords and entities with a transparent public/synthetic model source.
Breast Cancer WDBC Classifier
Public Wisconsin Diagnostic data via scikit-learn. Educational only — not clinical diagnosis.
Colorectal Cancer Risk Lab
UCI SUPPORT2 colon-cancer subset; an educational risk model with explicit limitations.
Enterprise Readiness Engine
Regression, anomaly detection and strongest/priority dimension analysis.
Smart-Home Telemetry Intelligence
Occupancy probability, telemetry anomaly detection and automation recommendation.
Enterprise Operations Risk Intelligence
Operational risk probability, anomaly signal and explainable drivers without client records.
Retail CRM Intelligence
RFM-style segmentation, churn probability and next-best-action recommendation.
Metrics come from actual server training artifacts.
Real data where public. Synthetic data where privacy matters.
That distinction is intentional engineering, not a limitation hidden from the viewer.