Utilized Data collection strategies (from social network and internet and Public APIs), Natural language processing, Time series forecasting to build A fully automated market research platform in Fashion /Ecommerce industry to help online retailers, designers make better product selection, make better inventory management. Tech stack: Python,nltk, Scikit-learn, Prophet,Statsmodels, Aws Dynamodb, S3, Aws lambda functions,AWS API gateway, AWS cloud Watch. Designed a robust product roadmap , from research and technical proposal until deployment into Production.Utilized robust forecasting Machine learning algorithms to forecast hidden trends in fashion ecommerce.Utilized recommendation algorithm to provide customers with untapped products based on forecasted trends.

-Providing AI/ML based decision making solutions to address procurement and budget planning challenges. -Delivered a full quantity recommendation system that contains 3 fundamental forecasting pipeline to support business in efficient buying decision and minimize overbuy and underbuy. -Delivered fundamental components of Auto replenishment system, to forecast demand and lead days.-Providing AI based solution for Logistic and Operation - Delivered an ML based system to efficiently allocate right amount of products in different Fulfilment Centres and transport the rest with effecient transportation method to reduce the cost and increase speed to the market. -Delivered an END to END inbound forecasting complex solution that helps business to forecast volume and receipt date of each product across product journey from order date and ASN to receipt at inventory.-Implemented and end to end ETL to migrate google sheets into S3 bucket to minimize the risk of data loss or corruption.

Led a team of data scientists and Engineers to develop a real-time price prediction application.Extracted, transformed, and analyzed complex, high-volume data from different sources.Analyzed, optimized, and deployed Bond Market Issuance propensity machine learning pipeline with a highly imbalanced dataset.Analyzed, developed and deployed to production a recommender algorithm to recommend products to the potential investors.Leveraged descriptive and inferential analysis to analyze the liquidity risk of the securities.Leveraged descriptive and inferential analysis to perform customer segmentation analysis.Developed, configured, and queried an Amazon Redshift data warehouse and analyzed data using SQL Client to provide actionable insights from large amount of data.

Analyzed, built, tuned and deployed an accurate predictive model to analyze Marketing Campaign using Exploratory Data Analysis, Decision Trees, Logistic Regression, and SVM. Python, SQL, Sklearn library and Keras were the main technologies used in test and development.Explored, built, tuned and prototyped a Fraud detector pipeline and successfully achieved a reasonable F1 Score in a highly imbalanced dataset utilizing Applied Statistics, Machine Learning Algorithms such as Random Forest, SVC and KNN Classifiers and NLP techniques. Python, SQL, Flask framework Sklearn library and Keras were the main technologies used in test and development.Performed Customer Segmentation and Churn Analysis

Researched, developed, and tested advanced statistical algorithms (Polynomial Chaos, Stochastic Testing, Monte Carlo) to efficiently quantify uncertainties in radio frequency systems using Python, Pandas, SQL, MATLAB, Matplotlib in Ubuntu Linux environment. Performed descriptive and inferential multivariate statistical analysis to get insights and compare new Machine Learning algorithms with the traditional approaches. The results were within 95% accuracy.