Junior Data Scientist
2024 – PresentThe Digital Glow (Remote, contract)
Two-year remote engagement delivering machine-learning fault detection and BI across a solar and tech company's installed solar fleet, covering model development, data pipelines, explainability and ongoing monitoring.
- Built a two-stage fault-detection system in Python (an XGBoost baseline on engineered telemetry features, followed by an autoencoder for anomaly detection), monitoring 200 live solar systems across inverter, battery and panel data.
- Raised the fault-catch rate from 78% to 86% and cut average downtime per fault from 3.5 hours to 2 by surfacing faults earlier for proactive maintenance.
- Engineered SQL pipelines aggregating multi-source telemetry and CRM data, and applied SHAP to surface the key underperformance and fault drivers for the operations team.
- Built Power BI dashboards for BAU monitoring of energy production, system health and commercial KPIs, and monitored deployed models for performance drift over time.