Soban Zamir

•  Developed a physics-informed deep learning pipeline (PyTorch) using Transfer Learning to predict real time EV energy consumption for new vehicle models with zero historical driving data. •  Engineered a sequential base model (GRU/LSTM) trained on the open source VED (Vehicle Energy Dataset) to capture macro driving patterns and ambient thermal impacts. •  Implemented Few Shot Learning and layer-freezing techniques to adapt the pre trained model to a target vehicle using fewer than 5 trips of data. •  Ach

• Assisted senior engineers in preparing and maintaining technical documentation for electrical projects. • Reviewed electrical drawings, wiring diagrams, and project specifications to support project documentation and compliance. • Maintained project records, equipment documentation, and installation reports to ensure accurate project tracking.

• Assisted in the installation and configuration of IoT based temperature monitoring sensors used in industrial refrigeration and storage environments. • Supported testing, calibration, and troubleshooting of temperature sensors and communication gateways to ensure accurate data transmission. • Monitored real time temperature data via monitoring dashboards and helped prepare technical documentation and installation reports.

Reduced cost per lead by 18% through Facebook Ads optimization and audience segmentation. Boosted engagement rates by 40% by developing data-backed ad creatives using performance analytics. Automated weekly campaign reports using Excel, saving 5+ hours/week.

Gained hands on skills in SQL, R, Tableau, Excel, and data storytelling through real world projects, learning to turn raw data into actionable insights