Gursimran Singh Email and Phone Number

• Engineered a cutting-edge object detection pipeline to achieve 100% accuracy in detecting anomalies while innovatively expanding the image generator module for enriched analytical functionality.• Pioneered an ensemble method integrating AlexNet, ResNet, VGG, and custom CNNs, which propelled anomaly detection accuracy to 99% and significantly reduced false positives, enhancing system robustness without additional training overhead.

• Developed and Dockerized a full-stack application with Flask and Vue.js, consolidated diverse log sources, enabled Jira ticket creation, and enhanced issue detection with NLP-powered text preprocessing.• Efficiently reduced debugging time by 30%, leading to improved troubleshooting efficiency and accelerated problem resolution, and deployed RESTful APIs for features like redundancy removal and comparison across sources.• Utilized LLM, BERT, and GPT-4 to innovate data-driven… Show more • Developed and Dockerized a full-stack application with Flask and Vue.js, consolidated diverse log sources, enabled Jira ticket creation, and enhanced issue detection with NLP-powered text preprocessing.• Efficiently reduced debugging time by 30%, leading to improved troubleshooting efficiency and accelerated problem resolution, and deployed RESTful APIs for features like redundancy removal and comparison across sources.• Utilized LLM, BERT, and GPT-4 to innovate data-driven decision-making and custom chatbots and patented the method. Show less

• Designed and implemented a system for solving the distribution center-to-store alignment problem by creating an algorithm that predicted the amount of inventory that should be pushed from the distribution center to the stores.• Improved model accuracy by 32 times over the prior model by implementing a two-phase ML pipeline that generates embeddings using neural networks and feeds them to tree-based models (CatBoost, XGBoost) and engineered large warehouse data solution.

• Developed complex queries and large-volume data pipelines for audience selection strategies for a Fortune 25 retail giant.• Leveraged ETL processes for data integration, which drove statistical models, boosted sales by 12%, enabled data-driven promotional campaigns, and utilized NLP for customer segmentation.

• Developed an IoT-based device using devices and sensors to collect real-time data from fields to facilitate crop-watering decisions• Analyzed collected data and orchestrated machine learning pipelines using techniques like PCA, Gradient Boosting Regression Trees, SVM, SVR, multiple linear regression, and K-means clustering to predict future soil moisture with an accuracy of 93%