Karthik Reddy

Owned the full AI product lifecycle for hospital discharge workflows on 100,000+ EHR records, building readmission risk scoring deployed via Flask APIs and AWS. Designed an LLM + LangChain clinician interface and translated SHAP explainability into clinician-friendly risk summaries, reducing discharge time-to-decision from minutes to seconds with production MLOps using DVC, Docker, and CI/CD.

Defined and delivered a full data strategy for a regional UK retail chain, using market basket analysis and segmentation to drive revenue decisions. Built XGBoost + Prophet forecasting and Tableau dashboards for store managers, achieving a 12% upsell revenue increase through improved product placement and cross-promotional bundles.

Owned delivery of a real-time anomaly detection platform for live SCADA sensor streams, shifting operations from reactive breakdown response to proactive monitoring. Reduced false alerts by 40% and cut operational costs by 25% using Isolation Forest and autoencoders, integrating geolocation and weather APIs and delivering actionable Plotly Dash and Power BI dashboards.

Built a predictive maintenance platform for a major industrial facility monitoring 120+ machines in real time. Developed Python ETL for IoT sensor ingestion, trained LSTM and Random Forest models to forecast breakdowns with 85%+ accuracy, and operationalized outputs through live Dash + Plotly dashboards with mechanical engineers to reduce unplanned downtime.

Completed a Bachelor of Technology in Electrical and Electronics Engineering at Vignan’s University.

Completed a Master of Science in Data Science at the University of Greenwich in London.