Dipan Ghosh

About Dipan Ghosh

I am a Data Scientist with hands-on experience in EV energy, pharma, and digital media industries, delivering predictive models, time series forecasts, and marketing analytics solutions that drive measurable business impact. My expertise lies in building scalable ML systems, automating pipelines, and applying advanced statistical methods to solve real-world problems.In the EV sector, I designed and deployed predictive maintenance, e.g. uptime, utilization, downtime, reliability and revenue forecasting models for charging stations using XGBoost, ARIMA/SARIMAX, and AWS SageMaker, built automated ETL pipelines from Redshift, and applied advanced feature engineering to generate model-ready datasets. Collaborated on the early-stage design of a Generative AI chatbot (AWS Bedrock), assisting in prototyping workflows for customer service automation.In the digital media industry, I developed Unobserved Component Models (UCM) and Marketing Mix Models across TV, search, social, and email, automated ETL workflows (40% faster), built forecasting pipelines (ARIMA, ETS), and created dashboards that reduced reporting turnaround by 50%. I also contributed to rebranded product modeling and cannibalization analysis to measure cross-product effects.In the pharma sector, I built ROI optimization models using (OLS/WLS regression) to evaluate and optimize media spends, helping marketing teams improve allocation by 5–10%.Additionally, I created an ANCOVA-enabled Streamlit app for A/B testing and causal inference, empowering non-technical teams to run valid experiments. I have also mentored peers and interns on and regression approaches, ensuring knowledge transfer and statistically sound outputs.Beyond industry projects, I continue to strengthen my expertise with self-learning in credit risk modeling (WOE/IV, scorecards, PD, KS, Gini, ROC-AUC) and exploring GenAI applications, having had early exposure to AWS Bedrock chatbot development.Core Skills: Python, R, SQL, AWS SageMaker/Redshift, XGBoost, LightGBM, ARIMA/SARIMAX, UCM, Regression, Ridge, LASSO, Elastic Net, Feature Engineering, Causal Inference, A/B Testing, Data Pipeline Automation, Dashboarding.

Experience

1. Senior Analyst

   MMA (Marketing Management Analytics)

   Jan 2026 — Present · Bangalore, IN

Education

• Bidhannagar Government College

  Bachelor's degree, Statistics

• University of Calcutta

  Master's degree, Statistics