Ezequiel Cutin

• Ran autonomous-driving test operations in downtown Miami, validating behavior in complex urban traffic and pedestrian-heavy scenarios.
• Logged and reviewed telemetry to spot deviations, prioritize fixes, and support safe, repeatable iteration of driving behavior.
• Coordinated with engineers to capture edge cases and improve testing/validation coverage for real-world reliability.

Skills:

Autonomous Systems · Software Testing · Data Analysis · Technical Documentation · System Validation · Performance Monitoring · Quality Assurance

• Facilitated student understanding of entrepreneurial B2B sales and innovation in both undergraduate and graduate-level courses.
• Acted as an investor for student projects, evaluating business proposals and providing constructive feedback.
• Provided guidance on sales strategies, intellectual property, and investment acquisition, enhancing students' practical knowledge.
• Developed essential soft skills through teaching, presentation, and conflict resolution activities, contributing to a collaborative learning environment.

Skills:

People Management · Entrepreneurship · Evaluating Projects · Mentoring · Sales Strategies · IP (Intellectual Property) · Investment Acquisition · Conflict Resolution · B2B Sales · Innovation

• Supported an AI-assisted hiring workflow in a SaaS product, helping generate structured interview prompts and improve consistency in candidate review.
• Participated in scrum and UX collaboration, then ran beta user interviews to gather feedback and guide iteration.
• Built automated checks with testim.io, including validation for CAPTCHA verification flows, to improve reliability during releases.

Skills:

Product Management · Scrum · Test Automation · Competitive Analysis · GenAI · Agile Methodologies · Testim.io · React.js

• Participated in an intensive 3-week startup accelerator hosted by the European Innovation Academy (EIA), focused on building and launching entrepreneurial ventures.
• Co-founded Ugly Buckets, a startup blending fashion and personal safety, collaborating with an international team of 5 members.
• Took the lead on all technical aspects of the project, including building a responsive website, prototyping 3D product models, and setting up the backend for e-commerce functionality.
• Worked closely with mentors from the tech and business sectors, culminating in a final pitch to a panel of investors.

Skills:

Entrepreneurship · Market Research · Scrum · 3D Modeling · Startup Development · Pitching to Investors

Images:

❮ 1 of 3 ❯

• Conducted in-depth analysis and validation of housing imagery to ensure compliance with Detroit's Census criteria for identifying habitable residences.
• Utilized advanced geospatial data analysis techniques to assess housing conditions and accuracy of imagery data.
• Coordinated with cross-functional teams to integrate housing data into the city's comprehensive database, ensuring alignment with Census data for precise fund allocation.
• Applied machine learning algorithms to predict housing habitability, enhancing the city's data-driven decision-making process.
• Developed detailed reports and visualizations to communicate findings and recommendations to city officials, facilitating informed policy-making.

Skills:

Data Analysis · Geospatial Data Analysis · Machine Learning · Data Validation · Report Generation · Cross-Functional Collaboration · Policy Recommendations · Advanced Excel · Python · SQL

Spearheaded the development of a highly efficient and scalable customer support microservice using Node.js, Express.js, and MongoDB. Designed and implemented secure authentication and role-based access control to ensure data security and compliance. Developed comprehensive CRUD operations and integrated real-time notifications to enhance user experience and operational efficiency. Successfully deployed the microservice on Heroku, complemented by detailed Swagger API documentation for seamless integration and collaboration.

Skills: Secure Authentication · Role-Based Access Control · Real-Time Notifications · API Creation

Images:

This project focuses on developing a robust search engine using advanced algorithms and tools to optimize document retrieval. The search engine integrates Hadoop MapReduce, TF-IDF scoring, and Google's PageRank algorithm to enhance search accuracy and efficiency. The system is deployed on AWS to ensure scalability and performance.

Skills: Document Retrieval · Algorithm Implementation · Full-Stack Development · Cloud Deployment

Disclaimer: This repository is private due to the University of Michigan honor code policies. It was developed as part of the EECS 485 course (Web Systems) at the University of Michigan. Please contact me directly if you need more information about this project.

• Developed a fully functional MapReduce server for the EECS485 web systems class, utilizing Python to handle large-scale data processing tasks.
• Implemented advanced networking techniques using TCP and UDP protocols to manage communication between server components, ensuring efficient data transfer and task coordination.
• Enhanced server performance and responsiveness through multithreading, enabling concurrent processing of multiple tasks.
• Designed and developed a Master and Worker architecture to effectively distribute workloads, manage task execution, and ensure system scalability.
• Incorporated fault tolerance mechanisms to detect and recover from worker failures, maintaining the reliability and stability of the server.
• Enabled efficient inter-worker communication to coordinate task execution and share intermediate results.

Skills:

Fault Tolerance · Multithreaded Development · Git · Network Programming · Multithreading · TCP · MapReduce · Data Processing · WebSocket · Systems Design · Concurrent Programming · Python · UDP · Distributed Computing

Disclaimer: This repository is private due to the University of Michigan honor code policies. It was developed as part of the EECS485 course (Web Systems) at the University of Michigan. Please contact me directly if you need more information about this project.

• Insta485 is a set of 3 projects completed in UMich's EECS485 – Web Systems class. These projects focus on full-stack development of an Instagram-like system, aiming to incorporate all the basic user flows of an image-sharing site.
• The frontend was built using React, Jinja, HTML, and CSS, and served using Python's Flask framework.
• The backend consisted of a REST API built with Python's Flask and a SQLite database.
• Deployment was carried out on AWS using tools like EC2, ensuring robust and scalable infrastructure.

Skills:

HTTP · REST APIs · Full-Stack Development · WebSocket · JavaScript · Python · Jinja

Technologies Used:

Disclaimer: This repository is private due to the University of Michigan honor code policies. It was developed as part of the EECS 485 course (Web Systems) at the University of Michigan. Please contact me directly if you need more information about this project.

Video:

This project is dedicated to the development of an advanced AI model that excels at detecting architectural styles. The model has undergone rigorous training on a diverse dataset consisting of 10,000 high-quality images encompassing 25 different architectural styles. These images have been carefully selected from the renowned Kaggle architectural style database (accessible at Kaggle Database).

Skills: Computer Vision · AI Model Creation · Model Training · Python · Machine Learning

Developed a dynamic web application, DinoSoar, that aggregates and presents detailed information on European flights. Leveraged Node.js and key airline APIs to seamlessly integrate real-time flight data and pricing models. The application was designed with a focus on optimizing accessibility and enhancing user experience, providing users with an intuitive and efficient platform to access comprehensive flight information.

Skills: Web Development · API Integration · User Experience Design · Real-time Data Processing · Full Stack Development

Developed a generative AI model using diffusion methods to produce lifelike floral images. Improved model efficacy by experimenting with various weight initializations and activation functions, and optimized the model's architectural depth to enhance adaptability and image quality.

Skills: Generative AI · Diffusion Methods · Model Optimization · Python · Machine Learning

Images:

Analyzed customer data to predict personal loan uptake, leading to optimized marketing strategies. Processed and organized a dataset of 5,000 customer records using SQL for accurate analysis. Created a logistic regression model in Python, achieving 92.16% accuracy, to guide targeted marketing towards customers with CD accounts and advanced education.