Prashant Kaundal

Leading a team of 12+ Decision Scientists, managing a $1M+ engagement portfolio for a Fortune 50 US Home Improvement retailer. Overseeing day-to-day operations, project delivery, and client interactions; driving strategic planning, data analysis; and managing multiple threads to support business objectives.

Text Analytics

• Developed and implemented a framework using large language models (LLMs) to analyze customer survey data. Applied sentiment analysis, topic modeling, embeddings, and cosine similarity to deliver insights.
• Identified pain points and created strategies that improved LTSA by 13%, aiding leadership decisions.

Customer Lifetime Value

• Pitched and implemented the Bayesian BG-NBD model to evaluate customer lifetime value in the DIY segment. Achieving a validation R² of 0.7 over one year, the model advanced business intelligence, enabling the marketing team to target high-value segments and enhance campaign effectiveness.

Worked as a Product Analyst for a tech giant. Led a team of 4 to analyze customer behavior, providing data-driven insights for product optimization and acquisition growth. Focused on enhancing customer engagement through strategic recommendations and served as the primary communication point with clients to align on business objectives.

Anomaly Detection

• Identified suspicious behavior on an established communication platform for video calls. Implemented the Isolation Forest anomaly detection model to flag fraudulent users, preventing an estimated 18% in potential revenue loss due to fraud.

Customer Segmentation

• Helping the client gain a clearer view of their customer landscape (video discussion platform), we conducted clustering analysis and segmented customers into five groups based on KPIs and behavioral metrics. This led to targeted feature enhancements and personalized marketing, boosting engagement rates by 20%.

Product Analytics

• Developed reports and visualizations using Power BI to provide insights into user engagement and growth for the employee experience platform during its growth stage. Data-driven findings offered actionable recommendations for feature development and customer retention, enhancing user experience and platform success.

Market Mix Modelling

• Developed a Bayesian Market Mix Model (MMM) to optimize marketing spend, collaborating on data preprocessing, feature engineering, and model building with Python and PyMC3. Conducted MCMC sampling and validation, integrating digital, TV, and offline data, leading to a 25% improvement in ROI