Lead Data Scientist
Zee
Total years of experience :5 years, 6 Months
As a part of the DS Product Analytics in Zee, leading all data science initiatives for two charters to develop data driven solutions which would help in taking key business decisions for one of top OTT platform with 1Billion users i.e., Zee5 OTT
Customer Targeting:
Using supervised ML models created propensity models which helped in targeting new users for multiple new features.
§ Lead a team in developing and deploying multiple propensity models for AVOD marketing team, for targeting In-App game users and ILT20 sports viewers. Resulting a notable increase of user engagement by 25%.
§ Implemented cohort-based targeting for advance renewals and lapsed users, resulting in a revenue increase of 6 Cr during FY23-24, driving sustainable growth and maximizing customer lifetime value.
§ Achieveda2%increaseinviewerretentionbyimplementingapredictivechurnmodelforAVODviewersandleveraging content recommendations through CLM intervention.
Device Intelligence and Causal Relationship:
Created a framework to identify RCA’s and web scrapping to identify user device meta data for personalization.
§ Developed a Flask API and data table powered by web-scraped device metadata, enabling personalized targeting, enhancing ML model performance, resulted in saving ~1.5 Cr by eliminating external dependencies.
§ OrchestratedZee5UsertrackingdatamigrationtoGCP.ImplementedcausalregressionmodelidentifyingRCAsacross user journey, enabling proactive email alerts that saved 100+ manual hours which helped in arresting KPI drops.
Gen AI and DIH:
Created a web-based UI which showcases the key metric for the health of Zee5 App to leadership of Zee.
§ Deployed DIH (Data intelligence hub) an web-based UI on Azure App services, showcasing Zee5 app health and enabling alert tracking for key KPIs across 9 charters for PMs and CEO's office.
§ Developed an MVP using Gen AI Lang chain models which converts share the data from text prompt which helps business team to access basic data without any delay, thereby saving the turn-around time to less than a min from hours.
stakeholders
▪ Experienced working within a team and managing a small one.
▪ Machine Learning Algorithms: Linear Regression, Logistic Regression, Decision Tree, Random Forest, GBM, K-
means, SVM, KNN, PCA, Time Series, Sentimental Analysis, probabilistic models
▪ Languages: SQL, R, Python.
▪ Visualization: Ploty Dash, Tableau, Excel, Google Analytics.
▪ Others: Hadoop, PySpark, TensorFlow, Jupyter, Github, BizTalk
As a part of the Marketing Analytics and Global sales services team in Affine, I am involved in providing data science
expertise and insights for the world’s largest Gaming platform i.e., Sony PlayStation
Forecasting and Adjusting Game Sales
To identify the key sales drivers of First Party Games and their impact on sales. Forecast the sales generated by
the game using identified drivers
▪ Created a regression model for forecasting Sales of upcoming major games with MAPE less than 15%.
▪ Used Time Series in forecasting sales of franchise games which further helped in deciding partnership with
Studios.
Install Base Management
Forecasting sales of console for inventory management.
▪ Random Forest/Regression to predict the sales of console for the next finical year, starting from variable selection
to delivering the final model to client.
▪ Reduced the MAPE of the existing model by incorporating new variables and using various other algorithms.
Assigned data scientist for Onsite business team
Continuously providing support to the onsite business team and being the primary contact in giving data driven insights
Other Projects: Used disc analysis, Forecast dashboard, Strategic value targeting
,
,
,