Eshaan Minocha
Machine Learning Engineer ✉️ Email | 🔗 LinkedIn | 🐙 GitHub
I am a passionate Machine Learning Engineer with a strong foundation in both software engineering and advanced data science. My expertise spans the end-to-end development and deployment of machine learning solutions, from designing robust data pipelines to building and optimizing predictive models for real-world impact.
With hands-on experience at The Goodyear Tire and Rubber Company, I have led initiatives to enhance model reliability, integrate uncertainty quantification, and drive cost-saving innovations by leveraging open-source optimization tools. My work has directly contributed to improved forecasting accuracy, operational efficiency, and strategic decision-making in a Fortune 500 environment.
I am skilled in a wide array of technologies, including Python, SQL, cloud platforms (AWS, GCP), and modern ML libraries such as PyTorch, XGBoost, and TensorFlow. My projects range from developing scalable machine learning pipelines and real-time data streaming modules to architecting trustworthy registries for software modules and benchmarking robust deep learning techniques.
I thrive at the intersection of research and engineering, always seeking to bridge the gap between cutting-edge algorithms and scalable, production-ready systems. I am deeply interested in Bayesian modeling, time series forecasting, optimization, and the trustworthy deployment of AI in industry.
Currently, I am open to full-time opportunities in Machine Learning, Data Science, or AI Engineering, and I am eager to collaborate on innovative projects that push the boundaries of what data-driven technology can achieve.
Experience
-
Data Scientist @ The Goodyear Tire and Rubber Company
Akron, OH • June 2024 - Present
I developed scalable machine learning pipelines and integrated uncertainty quantification techniques, enhancing model reliability and predictive performance. I led cost-saving initiatives by replacing proprietary optimization tools with open-source solvers, reducing annual expenses by $40K. My work on Gaussian Processes and multivariate time series modeling significantly improved inference efficiency and forecasting accuracy, supporting strategic demand planning and operational decision-making. -
Data Science Intern @ The Goodyear Tire and Rubber Company
Akron, OH • May 2022 - Aug 2023
I migrated ML applications to the cloud, reducing downtime to just 0.1%. I managed and optimized AWS DynamoDB databases for efficient ML payload storage and retrieval, cutting query complexity by 50% through smart schema design. Additionally, I benchmarked outlier detection techniques to automate anomaly detection in ML pipelines, enhancing system reliability. Include any relevant technologies, methodologies, or projects you worked on. -
Software Engineering Intern @ Hughes Network Systems
Gaithersburg, MD • May 2022 - Aug 2022
I built a low-latency real-time data streaming module using WebSockets (~1 ms latency) and optimized JSON structures for asynchronous backend communication. Deployed an Express-based REST API powering real-time analytics dashboards for system performance monitoring.
Machine Learning Projects
Gaussian Processes & Bayesian Optimization
Full-stack system for generating predictions using Gaussian Processes and performing Bayesian Optimization to optimize material features. Implemented advanced ML techniques for material science applications.
Learning to Reweight Examples for Robust Deep Learning
Investigated the vulnerabilities of deep neural networks to training set biases and noisy labels, improving outlier detection accuracy by 5% and successfully replicating label detection trends from the original research. Implemented and evaluated robust learning techniques to enhance model reliability.
Deep Image Prior
Modified convolutional neural network architectures for image inpainting and restoration using PyTorch, achieving 40% accuracy in restoration experiments. Applied advanced clustering and evaluation techniques to assess model performance on diverse image datasets.
Kaggle Competitions
Collection of Jupyter notebooks from various Kaggle competitions. Demonstrates expertise in data science competitions, feature engineering, and model optimization across diverse datasets.
The Subtle Art of Not Giving a Flix
Full-stack recommendation engine for books. Built a sophisticated system that provides personalized book recommendations using advanced recommendation algorithms and user preference analysis.
Software Engineering Projects
Trustworthy Registry of Modules
Developed a CLI and API system to assess the trustworthiness of npm packages, with core functionalities for upload, download, and update. Leveraged TypeScript, Python, and GCP Firebase to build and deploy a secure, cloud-based registry for package management.
Personal Website
This website! Designed and developed a modern, minimalistic, and fully responsive personal portfolio using HTML, CSS, and JavaScript. Features a dark theme, interactive skills sidebar, and smooth navigation for an engaging user experience.