Author: Mayukh Dattachaudhuri
-
Nucor Steel — Manufacturing Planning & Saleability Analytics
I worked on a manufacturing analytics project focused on improving planning decisions for non-standard steel coil transitions at Nucor Steel. The project analyzed how production, quality, demand, and variability factors influence the saleability and cost impact of different grade transitions within a steel manufacturing environment. I developed a data-driven penalty model to evaluate transition risk…
-
Autodesk E-Commerce Database Design: SQL & Business Reporting
I completed a database design project focused on improving data consistency, scalability, and reporting accessibility for Autodesk’s e-commerce operations. The project was based on a real-world business case involving digital sales, customer accounts, subscriptions, software products, transactions, pricing, usage records, and support tickets. I designed and refined an EER diagram and relational schema that mapped…
-
Walmart U.S. Retail: Inventory Forecasting & Safety Stock Analysis
I completed an inventory analytics project focused on improving retail replenishment decisions across Walmart-style SKU demand patterns. The project explored how demand variability, reorder points, safety stock levels, holding costs, and stockout risks interact in retail inventory planning. Using Python and Excel, I developed demand forecasting and safety stock models to evaluate reorder point decisions…
-
Reeview.ai: Human-Like AI Video Interviews for Recruiting
I am a co-founder of Reeview.ai, a startup idea for an AI recruiter capable of conducting real-time, human-like video interviews. Unlike traditional screening tools, Reeview.ai provides customizable, company-specific AI agents that hold natural conversations, surface KPIs and red flags, and integrate seamlessly with existing ATS and HR platforms. Our team won “Most Impressive Tech” at…
-
Predicting House Prices with Regression and Machine Learning
For my IEOR 165 project, I explored the relationship between housing features and sales prices using regression-based machine learning models. I implemented and compared Ordinary Least Squares (OLS), Ridge Regression, and Nadaraya-Watson Kernel Regression, training each model on an 80/20 data split and tuning hyperparameters through cross-validation. Among the three, OLS achieved the lowest error…
-
VersaCrush — An Inclusive and Sustainable Multi-Functional Device
I co-developed VersaCrush, a multi-functional compression tool designed to make daily tasks like crushing cans, compressing paper, and wringing clothes more accessible for people with limited mobility or strength. Our team combined ergonomic design, iterative prototyping, and user feedback to ensure the product was both practical and user-friendly. Beyond accessibility, VersaCrush also promotes environmental sustainability…
-
Travel Itinerary Optimization Model
For my IEOR 164: Optimization project, I developed a stochastic optimization model in Pyomo to design a cost-effective and enjoyable three-day trip to Zion National Park. The model incorporated uncertainties in lodging costs, flight prices, and travel durations while making decisions on transportation, lodging, dining, and activities. By applying scenario-based optimization, I generated an itinerary…
-
Optimizing Wind Turbine Design for Greater Efficiency
As part of UC Berkeley’s Engineering 26 course, I co-led a team project to design, fabricate, and test a miniature wind turbine system. We focused on optimizing blade geometry, tower stiffness, and aerodynamic efficiency using CAD modeling, 3D printing, and structural analysis. Our prototype incorporated a hybrid tower design inspired by jacketed caisson and tripod structures, paired with curved blades modeled after advanced airfoil…
-
EZ Release Hook- Lead Engineer
I served as the lead engineer for EZRelease, an innovative non-piercing fishhook designed to make recreational fishing more sustainable and humane. Unlike traditional hooks, EZRelease uses a spring-based mechanism that gently expands within the fish’s mouth, securing the catch without causing incisions or internal injuries. To validate our design, we conducted extensive user testing with both experienced anglers and novice…
Mayukh's Projects 