Lifease Solutions specializes in providing innovative data analytics and machine learning solutions, with a focus on sports analytics. They help businesses transform raw data into actionable insights, streamline data processing, develop predictive models, and create intuitive visualizations to support data-driven decision-making.

Implemented a machine learning prediction model using Logistic Regression and Random Forest Classifier to determine winning percentages of cricket teams, achieving over 85% accuracy.

Analyzed key cricket metrics for 20 teams across 24 years and for all three formats (Test, T20I, ODI), embedding filters like year range, over range, and bowler spell number.

Developed a data pipeline at Lifease Solutions to efficiently convert JSON files into CSV format, leveraging PySpark and Pandas for comprehensive analysis of 100GB of data.

Utilized advanced Python libraries (NumPy, Pandas, Scikit-learn) and data visualization tools (Matplotlib) to derive insights and present data effectively, creating an analysis dashboard for all matches.

Lincode.ai is a company that uses AI to help manufacturing industries inspect their products more efficiently. Their technology automates the process of checking for defects in products, ensuring high accuracy and faster production times. This helps manufacturers save money, reduce errors, and improve their overall quality control processes.

Developed anomaly detection models to identify defects in vehicle logos, leveraging machine learning techniques for accurate recognition and analysis.

Implemented image segmentation using the YOLO (You Only Look Once) algorithm to isolate and classify different components within automotive images.

Attirox is an AI-powered platform that changes the way fashion works in India by offering personalized style advice. Their technology helps people feel confident about their unique style. Attirox uses AI to make fashion more personal and enjoyable for everyone.

Developed a fashion recommendation engine at Attirox using Tensorflow and Keras leveraging ResNet50 for robust feature extraction and the K-Nearest Neighbors algorithm to recommend clothing items and combinations.

Implemented feature extraction and similarity matching by adapting the ResNet50 model, with a Global Max Pooling layer to convert images into feature vectors, enabling accurate and efficient nearest neighbor searches.

Improved user involvement by offering outfit ideas that meet individual tastes and the latest style trends, showing the tool’s ability to provide customized fashion advice.

Structured, practice-based, and outcome-oriented learning based platform in technologies like C++, Java, DBMS, OS.

Working as a TA for solving conceptual and query based doubts in Database Management System.

Solving 650+ doubs via chat & call support.

Achieved an average rating of 4.8 with 500+ ratings.