Durvesh Sainath Talekar

Accomplished Senior Data Analyst at Merkle with deep expertise in ML and Python; led 100+ market research projects, improved data accuracy and reporting speed, and saved 15 workdays/month through automation and scalable solutions; proficient in Quantum and SPSS, known for cross-functional collaboration that drives efficiency and impactful business insights.

E-commerce Customer Segmentation & Prediction - Machine Learning

• Designed an RFM-based segmentation model using Ward Linkage Hierarchical Clustering (Silhouette Score 0.76) to classify customers into key segments.
• Developed & deployed a Streamlit app to process raw transaction data, calculate RFM, and predict customer segments in real-time.
• Delivered actionable insights to enable targeted marketing and improve customer retention.

Loan Approval Prediction - Machine Learning

• Developed a machine learning pipeline that improved loan approval prediction accuracy to 96.75 % using a Random Forest model with optimized hyperparameters.
• Assessed performance using accuracy, confusion matrix, and classification report for robust evaluation.

IMDB Sentiment Analysis - NLP

• Developed a python pipeline to preprocess, analyze, review using NLP techniques to predict if a review is positive or negative.
• Preprocessed textual data through tokenization, stopword removal, and vectorization.
• Built and evaluated classification models to accurately predict positive or negative sentiments from text data.

• Client appreciations, Merkle, 12/23 - Received multiple client appreciations for delivering precise outputs on complex Pfizer, Google, Amazon, and JPMC projects with minimal intervention.
• Climb High Award, Merkle, 02/23 - Received the Climb High Award for consistently enhancing skills and ensuring on-time, error-free project deliveries.
• Make It Real Award, Merkle, 03/24 - Award for consistently exceeding expectations, exceptional work ethic, and multitasking ability.
• Key Contribution Award, Merkle, 06/25 - Honored for successfully managing high-priority accounts (Harris and Amazon) with 90%+ utilization and accountability.

• ENGLISH
• MARATHI
• HINDI