Machine Learning Engineer Intern

Thu Le

272-***-**** ******@********.*** Linkedin: /in/imbaothu/ GitHub: /imbaothu

EDUCATION

Bucknell University, College of Engineering, Lewisburg, PA Graduation: May 2026 Bachelor of Science, Computer Science

Relevant Coursework: Algorithms, Calculus, Cloud Computing, Compilers, Database Systems, Data Structures, Discrete Mathematics, Linear Algebra, Machine Structures, Operating Systems, Probability,Programming Languages, Web Development

● Two-Time Bronze Medalist, National Competition

● 36th place at 2024 ICPC Mid-Atlantic Regional Contest SPECIALIZED SKILLS

Languages: C++, Python, Java, JavaScript, SQL (PostgreSQL), HTML/CSS, R Frameworks: React, Next.js, FastAPI, Flask, Material-UI, Spring, Tailwind, Apollo Server, JavaFx Cloud Platform & DevOps &: Linux, AWS, Google Cloud Platform, Azure,Docker, Ansible, Jenkins, Terraform, TravisCI, Git Actions Databases & Data Warehousing: MongoDB, PostgreSQL, Redis AI/Machine Learning: TensorFlow, PyTorch, Hugging Face Transformers, scikit-learn, NLTK, spaCy EXPERIENCE

Undergraduate Research Assistant, Bucknell University, Lewisburg Aug.2024-Present

● Led research on AI-powered sEMG wristbands, analyzing 10+ existing designs to optimize motion detection accuracy

● Improved gesture recognition by by implementing spectrum analysis and ML models on 1000+ muscle signal samples for AR limb movement mapping

● Increased user satisfaction by through conducting studies with 50+ participants and implementing feedback for wristband ergonomics and response time

Artificial Intelligence Engineer Intern, VinBigdata, Hanoi, Vietnam May 2024 – Aug. 2024 VinBigData JSC drives digital transformation and efficiency for enterprises through advanced Big Data and AI solutions

● Achieved a 96% F1-score and 92% accuracy in a multi-label classification task for medical coding to accurately identify symptoms and correctly classify them with disease labels, improving diagnostic support.

● Applied advanced text augmentation techniques at both word and character levels to improve classification accuracy for rare labels.

● Boosted recall and precision on minority classes by 50% by implementing imbalanced data handling techniques such as MLSMOTE, MLROS, and MLRUS, significantly improving the model’s ability to predict rare medical codes.

● Improved precision-recall curve balance by 20% through experimenting with ensemble learning methods like HPSLPred and BR-IRUS, optimizing label correlations and reducing prediction bias across classes.

● Increased throughput in training by 30% and enhanced model generalization by utilizing cross-validation with stratified splits, ensuring stable performance across varying distributions of multi-label datasets.

● Processed and annotated 700+ medical images using polygon-based annotations, leading to a 15% improvement in the accuracy of computer vision models by refining input data consistency for multi-task learning approaches. Software Engineer Intern, Project X Vietnam, projectxvietnam.org, Remote Nov. 2023 – Dec. 2023 Project X is a non-profit nurturing Vietnam's next generation of tech talent

● Designed and deployed AWS Lambda functions for key backend operations, cutting server costs by 20% and ensuring high availability during peak traffic.

● Built and implemented a GraphQL API gateway using Apollo Server, reducing API response times by 40% and improving microservices communication.

● Refactored legacy PostgreSQL queries, implemented indexing, adopted partitioning strategies, reducing query execution times by 75% PROJECTS

BisonMessenger Java, JavaFX, GSON, WebSocket, JSON Nov. 2024 – Dec. 2024

● Collaborated in a 4-person team using Java, JavaFX, and GSON, enabling chat creation, messaging, and user search.

● Designed and implemented a scalable MVC architecture with distinct Controller, Model, and View packages

● Developed a file-based Database Manager with JSON serialization and synchronized collections for data persistence. E-Commerce Platform Next.js, FastAPI, PostgreSQL, Redis, MongoDB, Kubernetes, AWS Aug. 2024 – Oct. 2024

● Built a scalable microservices-based platform by deploying Docker and Kubernetes on AWS, ensuring high availability and efficient load distribution across services.

● Reduced query latency by 15% and boosted processing speed by 20% by implementing a robust data model using PostgreSQL and MongoDB, resulting in faster response times and improved user experience.

● Enhanced API performance and scalability by designing RESTful APIs with FastAPI, supporting high-throughput requests and seamless interaction for users.

Pinterest Clone React, Vite, Firebase, Tailwind CSS, JavaScript, Cloudinary June 2023 – July 2023

● Built a responsive Pinterest-style application using React and Vite, delivering a smooth, engaging user experience across devices.

● Enabled real-time data synchronization and authentication by integrating Firebase as the backend, ensuring seamless data updates and secure user interactions.

Contact this candidate