SAYYAPUREDDI VENKAT

• FULL STACK DATA SCIENCE, Almabetter, 02/01/23, 03/31/24
• CERTIFIED SYSTEM ARCHITECT, Pega, 01/01/21, 06/30/21
• DATA ANALYSIS BOOTCAMP, Trainity, 06/01/23, 01/31/24

INSTAGRAM USER ANALYTICS: Achieved 92% accuracy predicting viral Instagram trends using a multimodal XGBoost model trained on 1M+ posts with ResNet-50 features, spaCy NLP analysis, and user engagement data. Dashboard empowers content creators to optimize strategy for 30% higher engagement.

E-COMMERCE CHURN PREDICTION: Multi modal LSTM's predict e-commerce churn (Reduced by 12%). Stacked LSTM's analyze purchase history, sentiment from reviews, and demographics. Model achieves 0.88 AUC, surpassing traditional models and leading to targeted interventions that saved the platform significant revenue.

CLICKSTREAM ANALYSIS BOOST: An RNN model analyzes user browsing journeys to predict product recommendations in real-time. This approach led to a 20% increase in conversion rate, showcasing the power of personalized recommendations in e-commerce.

 DATA DRIVEN CUSTOMER LIFETIME VALUE OPTIMIZATION: This Power BI project transformed customer data into actionable insights, optimizing Customer Lifetime Value (CLTV) for a leading e-commerce platform. Utilizing RFM analysis, interactive dashboards, and potential machine learning models. This resulted in a 25% increase in average CLTV, significantly boosting revenue and customer retention.

PREDICTION OF COVID-19: Leveraged TensorFlow to develop a multi-model ensemble for COVID-19 classification using CT scans and X-rays. Employed four Convolutional Neural Network (CNN) architectures, achieving individual accuracies of 91%, 93%, 96%, and 89%.