Augmented Reality Virtual Try-On

1. Streamlined Augmented Reality (AR) Virtual Glasses and Watches Try-On using Unity 3D Glasses and Watches Models by overlaying on user's face and hand in real-time.

2. Developed low latency AR Virtual Try-On Android Application which contains a variety of Glasses and Watches Models to make it available for everyone to Try-On and buy.

Collected large data sets of more than 20 years of historic data of different lottery games that are played in different states in US. - Developed and enhanced statistical models by utilizing the best-in-class modeling techniques with available data. - Performed Exploratory Data Analysis (EDA), created data visualization graphics, translating complex data sets into comprehensive visual representations to find patterns and provided insights, progress and concise summary to peers and clients. - Analyzed Machine Learning algorithms like Naïve bayes, XGBoost and Random Forest to calculate most accurate winning numbers. - Developed and deployed this solution, tracked and monitored the performance of existing models, maintained and enhanced solutions over time.

Optimized and diagnose Natural Language Processing Pipelines to read, decipher and understand human languages by breaking the text into small chunks, reconstructs it, analyzes, and processes it to bring back the most relevant data from the Search Engine Result Page. - Hands-on experience on few problems like Named Entity Recognition, topic modeling, question answering, summarization, information retrieval/extraction, language generation/understanding, chatbots. - Work closely with other teams to develop system capabilities, models, controls, design data elements and implemented new features where the user can either choose from an existing event or search a new event and will get the summary of the event captured from multiple websites and recent tweets to calculates sentiment score, knowledge graph, and threat analysis. - Developed and deployed this solution, tracked and monitored the performance of existing models, maintained and enhanced solutions over time.

Researched on Computer Vision Technique that allows us to identify and locate objects in real-time, image and video. - Tracked and monitored the performance of existing Computer Vision Based Object Detection and Classification Models like YOLO, Faster RCNN, RNN and conduct review on the basis of Accuracy, Frames per second, and Time Taken to process the input image and give output. - Maintained and enhanced solution over time and streamlined it in Safety and Surveillance Domain which includes Vehicle Detection, Number Plate Detection, Person Detection, Face Detection, Personal Protective Equipment Detection of Construction Workers, and Anomaly Event Detection in real-time.

Created a Language Understanding System in Python which understand and interpret your sentences in textformat. - Explored new Information NLP sources like Google BERT, ALBERT, XLNet and ELMo to create predictive attributes that provide incremental discrimination in reading the sentences and predicting when the person is contradicting its own given sentences which can help in Legal Industry to identify and highlight when the accused person or the witness is lying during a legal proceeding.

- Solved a data science complex problem statement of identifying relationships between the items that people buy by looking for combinations of items that occur together frequently in transactions to make company sales grow by tracking and diagnosing market changes. - Ensured Modeling Accuracy and enhance modeling efficiency using Machine Learning Algorithms - Apriori and Fpgrowth to analyze and solve complex business problem of finding the relationships between the items that people buy. - Developed and deployed this recommendation solution which will help the retailers to drive billing, revenue growth and profitability through advanced analytical techniques and maximize business returns by institutionalizing efficient and accurate models through recommendation.