Kaiwen Gong

404-***-**** ad098x@r.postjobfree.com Atlanta, GA, 30308

https://github.com/Keviniscaiji https://www.linkedin.com/in/kaiwen-kevin-gong-a72043292/ PROFESSIONAL SUMMARY

Highly analytical and innovative-thinking Software Engineer pursuing Master of Science in Computational Science

& Engineering qualification with 5 years experience in utilizing various coding languages, collecting test results, and generating technical charts. Adept improving user experiences in systems, employing AI methodologies, and developing process automation models.

SKILLS

C, C++, Java, Python, Linux, Javascript, HTML, CSS, Node.js, Vue.js, jQuery, React.js, Git, GitHub, Flask, Spring Boot, Database, MySQL, SQLite, Android Development, Google Cloud, AWS, Machine Learning, Computer Vision. PROFESSIONAL EXPERIENCE

Research Assistant, Peking University 07/2021 - 11/2022

● Image preprocessing was carried out. Using OpenCV, image processing and cropping were implemented, leading to a 10% improvement in recognition accuracy during the experiment.

● An advanced Automatic License Plate Recognition (ALPR) system was developed by utilizing image processing algorithms, resulting in a 99.9% accuracy rate for extracting license plate numbers from vehicles passing through a location.

Research Assistant, University College Dublin 01/2022- 07/2023

● Introduced SYGNet, combining feature extraction, SVD-YOLO GhostNet modules, and VoxelNet-based algorithms to enhance autonomous driving perception in complex conditions.

● Achieved detection rates of 94.1% for cars, 90.9% for people, and 93% for edges, validating its efficiency by integrating YOLOv3 with GhostNet for optimized feature mapping on public datasets. PROJECTS

AI-powered online car sales platform, 02/2023 - 07/2023 Software Developer HTML, CSS, JavaScript, Python, Flask, Vue.js, Mysql https://github.com/Keviniscaiji/An-AI-powered-online-car-sales-platform

● Users can upload car photos, and the system, using machine learning, identifies the cars and provides automotive product links with an accuracy of over 90%.

● With Three.js and Vue.js, users can instantly render a car's paint job on-site and see the results in less than 2 seconds, significantly enhancing the user experience.

● Utilizing the FastText NLP model, the platform offers basic question-driven information retrieval, enabling the website to automatically respond to users' questions. FoodieKevin: Improved Maps Android App 08/2021 - 12/2021 Software Developer Android, Java, API, XML, Gradle, SQLite https://github.com/Keviniscaiji/FoodieKevin

● Through integrating Baidu map API with Android Studio, we implemented a feature not found in standard map software, allowing users to share their travel plans and receive restaurant recommendations along the route. Also, allow users to make common on restaurants along the route EDUCATION

Master of Science, Computational Science and Engineering [Expected] 08/2023 - 5/2025 Georgia Institute of Technology(Gatech)

Bachelor of Science, Computer Science 07/2019 - 07/2023 University College Dublin(UCDublin)