Vishal Bajetha

Customer Value Management (CVM)

• Responsible for building a customer value management (CVM) recommendation engine for telecom client that recommends offers based on customer behaviour and preferences.
• Designed and developed various machine learning models using Python, PySpark, along with heuristic-based use cases to understand customer behavior.
• Contributing in designing methodologies and implementing them for upselling, mitigating churn, reviving dormant accounts and cross selling, leveraging behavioral insights to drive revenue growth and increase user engagement.
• Oversaw on premise big data workloads for more than 50 million subscribers, utilizing SQL based technologies like presto, hive and MySQL.
• Monitor the performance of recommendation system in production and implementing improvements to enhance user engagement and experience.
• Automate and optimize the scripts for their fast and smooth run.
• Collaborating with the data engineering and UI/UX team to translate business requirements into scalable and effective solutions.
• Continuously analyzed post-production KPIs to assess the effectiveness of our recommendation system.
• Guided and supported team members and also mentored new team members in their understanding of project and data science methodologies.

PepsiCo Sales Forecasting

• Responsible for data analysis and model building to forecast sales specific to district, as well as supermarket level.
• Finalized ML models over traditional time series models, as it was giving better results.
• Worked on Azure VM

Spectrum Evaluation Workspace

• It was a telecom network optimization that provides the number of sites to be installed and deployment cost based on voice and data demand, coverage target, and other scenarios.
• Responsible for a major part of development, which includes a visualization module, frontend and backend changes, and code integration using R and R shiny.
• Worked on making the application run on dynamic inputs, reducing manual intervention, optimizing run time, and improving ease of use.

• Responsible for improving the debt collections by optimizing operational efficiency.
• Assessing data quality by checking for data format, abnormalities, patterns ensuring data consistency and data completeness.
• Exploratory data analysis
• Early warning indicator for the non workable cases based on different parameters.
• Performance improvement by monitoring and analysing different KPIs.
• Also assisted in decision making process for feature creation, improvement and integration in web application.

• Project on Urban Transport Scheduling of Electric Buses.
• The aim is to prepare an optimized schedule of Electric bus public transport bus routes and to carry out the feasibility analysis and inferences based on some real data.