[Yaseen Khalil]|Computational Modeler & ML Systems Architect

> Designed an early-stage orchestration layer for AI agents to interact with core business platforms through one consistent interface. The intent was to reduce integration complexity so product teams could automate workflows without rebuilding connectors for each service.

> Led a research direction centered on a Semi-Tensor-Product-based Graph Neural Network for cancer signaling analysis. The architecture uses STP operators to preserve logical structure while enabling differentiable learning, then integrates stochastic simulation outputs to study attractor behavior and intervention sensitivity as biological network scale increases.

> Engineered a production-oriented telemetry pipeline architecture with Airflow 3 and Postgres to transform raw vehicle streams into reliable analytics datasets. The workflow emphasizes replay-safe ingestion, backfill flexibility, and strong data contracts so downstream teams can trust operational and trend reporting.

> Built and iterated on machine learning workflows for vehicle-health signal interpretation, combining sequence models and clustering to detect risk patterns before service disruption. This work translated high-volume telematics streams into practical maintenance intelligence that supports faster intervention planning.