Kishore kumar Avulakonda

At Meta, I lead data science for a fitness analytics platform, building predictive models for 2M+ daily active users that improved engagement predictions by 25%. I productionize wearable telemetry models and personalized training recommendations, achieving sub-100ms inference, 18% feature-adoption growth, and 12% lower churn.

Previously at TCS, I built demand forecasting, inventory optimization, and production data pipelines for an enterprise order-management platform. I use Python, SQL, Flask, Docker, cloud data platforms, and optimization methods to turn models into monitored systems that reduce costs and improve operational outcomes.