Shivam Pradhan

Implemented advanced customer segmentation using ML models, RFM models, cohort analysis integrating CRM tool braze, used for targeting & enhancing business growth. • Developed and fine-tuned machine learning models, including Random Forest, for predicting subscription purchases and identifying voucher hunters, and RFM Cohort models for customer segmentation and CLTV prediction. • Conducted a detailed annual review analysis for e-Scooter users using Braze and Snowflake for each user group

Increasing efficiency and reducing time of ticket categorization model by adding a Labelling data manually, making categorical decisions, auto labelling with label studio Retrained on pre trained a trainable Embedding layer of TensorFlow. Deploying model, data and code using Git, S3 and DVC. Contributor in developing a ticket queuing system which assign ticket to the best service engineer based on SLA time and wait time.helping saving SLA Breaches. System designing of Feedback system for the ops queue

Creating QnA Chatbot using Aws Lex and deploying it on EC2 using Putty, Docker, Postman • Creating Chatbot using Aws Lex with SDK boto3 Integrating to Website using Cognito • Collecting data using Pytrends, Analyzing and predicting sales using Arima models • Used forecast library for finding right model and best values of parameters. • Forecasting results and visualize them. Pear Review and Presentation of the analysis • Created a POC on AB testing and Bandit Algorithms Compared different bandit algorithms