Pragyesh Mishra

Client- US Based Retail: Databricks Migration (MLOPS) - Setup ML FLOW Databricks for ML Model creation, training, tracking runs and experiments for hyperparameter tuning and deployment. Manual implementation of Databricks pipeline to deploy it as Kubernetes pods by containerization using Docker and hitting an End point API using Flask. Automated the process as CI-CD by using Terraform scripts and Circle-CI. Setting up a Central Model Registry. Client-US Based Ecommerce: Lift Calculation of Stores - Calculated Lift to understand if the closure of stores resulted in any growth in the online traffic and sales of nearby control stores. Client-Tredence Analytics (Under Progress): Sentiment Analysis - Created a Deep Learning NLP Product for Sentiment Analytics to understand market trends and end-user response.

Client–Australian: Churn Prediction Model Development - Defined Churn definition, performed EDA, trained a classification model for predicting propensity of churning, and performed cross-validation. Client–Australian: Recharge Likelihood Model Development - Defined recharge definition, explored customer profiles, trained a classification model for predicting propensity of recharging, and performed cross-validation. Client–Indonesia Based: Image Classification, Face Recognition - Collected 6000 color images from 60 employees. Performed image preprocessing and trained a CNN image classification algorithm, deploying it to the camera. Client –UK Based: Worked on customer segmentation and Campaign evaluation for Direct Mailings using Python, KNN, and Clustering. Projects included Customer Segmentation for Clothing, Baby, and Vegan products.

Handled the complete Recruitment Life Cycle, understanding client requirements, sourcing talent from online databases and social media sites, and coordinating interviews.