Projects:1. NeoMetry 📦🚀Description:NeoMetry automates the profiling and sorting of inbound shipments using high-speed conveyor lines. It accurately catalogs attributes like dimensions 📏, weight ⚖️, damage detection 🔍, SKU classification 🏷️, barcodes 📇, text 📝, and color 🎨 to boost logistics efficiency and order accuracy.Contributions:Led development of the profiling and sorting system 🛠️.Implemented real-time analysis algorithms and machine learning… Show more Projects:1. NeoMetry 📦🚀Description:NeoMetry automates the profiling and sorting of inbound shipments using high-speed conveyor lines. It accurately catalogs attributes like dimensions 📏, weight ⚖️, damage detection 🔍, SKU classification 🏷️, barcodes 📇, text 📝, and color 🎨 to boost logistics efficiency and order accuracy.Contributions:Led development of the profiling and sorting system 🛠️.Implemented real-time analysis algorithms and machine learning models 🤖.Customized the system to meet client requirements 🎯.2. NeoDIShA 🛍️📊Description:NeoDIShA is an AI platform for retail store owners with two applications:NeoDIShA Mobile App 📱: Allows store employees to capture shelf images 🛒📸 for real-time product analysis.NeoDIShA Web App 💻: Offers visual insights into sales 💰, product demand 📈, replenishment rates 🔄, and more.Contributions:Led development of Mobile and Web Apps for image analysis and data presentation 🌐.Implemented real-time analysis algorithms and machine learning models 🤖.Customized the platform to meet client needs 🎯.Successfully deployed in multiple Reliance Stores 🏬, enhancing inventory management and customer satisfaction 😊.3. Sentinel 🛡️🔍Description:Sentinel is an advanced AI-powered surveillance system designed to enhance security and monitoring. It utilizes computer vision and deep learning algorithms to detect anomalies, recognize objects, and provide real-time alerts 🚨.Contributions:Spearheaded the development of the Sentinel surveillance system 🚀.Integrated cutting-edge computer vision techniques for real-time anomaly detection 👁️‍🗨️.Collaborated with security teams to tailor the system to specific needs 🎯.Deployed Sentinel in high-security environments 🏢, significantly improving monitoring efficiency 🔒.Skills:Deep Learning 🤖 · Machine Learning 📚 · Project Management 🗂️ · Data Visualization 📊 · Team Leadership 👥 · Computer Vision 👁️ Show less

Projects:1. NIRS Modelling of Crude 🛢️🔬Description:Developed an ensemble machine learning pipeline to identify crude oil properties by analyzing absorbed Near-Infrared Spectrum (NIRS) data 🌈.Highlights:Ensemble Modelling 🎯: Architected a pipeline combining over-fitted and under-fitted ML algorithms for robust predictions.Data-Dependent Ensembling 📊: Enhanced crude property predictions by leveraging data dependencies.Cost Savings 💰: Achieved significant… Show more Projects:1. NIRS Modelling of Crude 🛢️🔬Description:Developed an ensemble machine learning pipeline to identify crude oil properties by analyzing absorbed Near-Infrared Spectrum (NIRS) data 🌈.Highlights:Ensemble Modelling 🎯: Architected a pipeline combining over-fitted and under-fitted ML algorithms for robust predictions.Data-Dependent Ensembling 📊: Enhanced crude property predictions by leveraging data dependencies.Cost Savings 💰: Achieved significant savings of over ₹1.72 crore per crude cargo through precise property predictions.Price Optimization 🤑: Implemented a crude purchasing price optimization algorithm to optimize procurement decisions.Impact:Revolutionized the industry with advanced prediction algorithms, achieving an MAE of 0.75 for accurate predictions and optimized purchasing decisions.2. Genomics 🧬📈Description:Developed an algorithm for gene sequence analysis to identify genotype-to-phenotype relationships in DNA 🧬.Key Outcomes:Efficient Analysis 🚀: Created a highly efficient gene sequence analysis algorithm, outperforming traditional methods.Optimized Efficiency ⏱️💾: Reduced time complexity from 225 minutes to just 12.2 minutes and space complexity from 21,504 MB to 160 MB.Improved Accuracy ✅: Overcame limitations of the traditional "P-value" method, achieving a 9.3% improvement with a phenotype prediction accuracy of 86.8%.Skills:Deep Learning 🤖 · Machine Learning 📚 · Data Visualization 📊 Show less

Projects:1. Labelled Dataset Generator Tool 🎥📝Description:Developed software to generate labeled datasets from video feeds recorded by vehicles 🚗📹.Achievements:Process Efficiency ⚡: Enhanced efficiency by ~80% through automated dataset generation.Impressive Accuracy 🎯: Achieved 92.8% accuracy at 30-35 FPS by training on the YOLOv2 model 🤖.Effective Object Detection 🚦🚘: The model effectively detects road signs and vehicles, aiding in autonomous… Show more Projects:1. Labelled Dataset Generator Tool 🎥📝Description:Developed software to generate labeled datasets from video feeds recorded by vehicles 🚗📹.Achievements:Process Efficiency ⚡: Enhanced efficiency by ~80% through automated dataset generation.Impressive Accuracy 🎯: Achieved 92.8% accuracy at 30-35 FPS by training on the YOLOv2 model 🤖.Effective Object Detection 🚦🚘: The model effectively detects road signs and vehicles, aiding in autonomous driving.2. Face & License Plates Blurring Tool 😎🔒Description:Designed a platform for detecting and blurring private information in vehicle data 📷🔍.Highlights:Efficient Platform 🚀: Reduced manual effort by ~97% while ensuring privacy protection 🛡️.High Accuracy 🎯: Achieved 88.7% accuracy at 8-10 FPS for face and license plate detection using the R-CNN model 🤖.Improved Blurring Efficiency 💨: Implemented optical flow tracking to handle missed detections, boosting blurring efficiency to 96.8%.Skills:Deep Learning 🤖 · Machine Learning 📚 · Data Visualization 📊 · Computer Vision 👁️ Show less