AWS based Web Service Development – Job Recommendation• Created Search, Favorite, Recommendation modules in Sprintboot Application containing content-based recommendation, session-based authentication, keyword extraction algorithms with RESTful APIs to handle 100k peak QPS HTTP requests and responses.• Designed and built relational database schema by MySQL on AWS RDS to persist, manage and process jobs data fetched from SERP API.• Incorporated OpenAI API to extract keywords from position descriptions.• Research and develop CRDT implementation to deal with distributed jobs like counters for streaming processing.• Setup and deploy the Springboot Application to AWS EC2 virtual machine instance operated by Linux command and installed Redis to cache Search and user favorite results, which reduced the latency 90% after testing.• Constructed UI components in the front-end using JavaScript, HTML, CSS, AJAX enabling users to search and apply for positions, add jobs to favorite as main factors on recommendation engine design.

• Conducted extensive research and reviewed academic papers to develop advanced point cloud and perception models, successfully improved the accuracy and reliability of autonomous driving systems.• Implemented deep learning frameworks PyTorch and MMDetection3D, creating demos to validate and showcase new algorithms in real-world applications, and deployed perception models for practical use, enhancing the efficiency of data processing tasks by 18%, crucial for real-time applications in autonomous vehicles.• Studied and implemented point cloud and image fusion models such as LSS, BEVFormer, BEVFusion, and SparseFusion to improve the integration of 3D data with 2D images, leading to more robust and accurate environment perception.

• Implemented enterprise applications of product information systems using Springboot, React, DynamoDB deployed on AWS.• Designed and developed systems receiving data from webpage with validation and persist in database monitored by Splunk.

• Developed machine learning models to improve robot functionalities such as navigation, manipulation, and interaction with the environment and humans. • Used PyTorch for model training, testing, and refining models based on performance and iterative testing, particularly useful in manipulation and navigation where models must adapt to new scenarios on the fly. • Utilized ROS tools and libraries for simulation and testing before deploying algorithms on physical robots. Conducted research for new machine learning techniques and methodologies that can be applied to robotic systems. Returned to the company in Jun, 2023 after a 9 month break.

• Robot Development: Design and construction of a sophisticated robot capable of precise ball handling – from collection at the bottom to accurate shooting at the top. Regularly documented and shared our team’s progress and milestones on social media, increasing community engagement and awareness. • Technical Seminars: Conducted extensive seminars across China, educating a wide audience of educators about the innovations and capabilities of our FRC team, thus spreading knowledge and fostering a wider understanding of robotics in education.• Organized weekly practices: Led weekly sessions, providing tailored programming guidance for coders and practical building instructions for engineers. Managed and updated the team’s official Twitter account, enhancing our digital presence by sharing insights into our weekly progress and business strategies.• Recruitment Initiatives: Successfully grew the SHSID FIRST Robotics Competition Team by over 20 members through persuasive school-wide speeches and interactive technology demonstrations, significantly boosting participation and interest in robotics at SHSID.

• Developed an AI-driven emotion recognition system using Python and TensorFlow, which accurately interprets emotional responses from visual cues with a 91% accuracy rate.• This project employed a convolutional neural network trained on a diverse dataset of facial images, over 10,000 labeled facial images, enhanced by real-time data augmentation and a custom triplet loss function to improve emotional state discrimination.• Research in Python, AI Lab Session, and Computer Vision

• In a team of ten, we accomplished AI and Data Science research project under the mentorship of ECUST professor until year-end, highly selective