Jiachen (Jason) Yan

- Designed and constructed an automated data ETL pipeline by integrating 10 million records from 16 different data metrics using PySpark, SQL and vendor API - Architected and delivered a software product that unified disparate datasets by retrieving information from 3+ sources and communicating with end users for demand analysis - Augmented the existing datasets by 8% with enhanced robustness after cross-referencing 3 data sources

- Established a Software-as-a-Service (SaaS) product that enables enterprise users’ code execution without physical hardware by utilizing TensorFlow API and XLA CPU workflow - Improved user experience of TensorFlow machine learning (ML) model application by integrating TensorFlow 2.x ML library with the enterprise software stack - Built and trained ML models (CNN, RNN, and BERT) that gave end users feedback on chips’ performance

- Performed machine learning research to break 6G bottlenecks at Huawei Canada, experimenting novel approaches to improve 5G technologies including Massive MIMO (multiple-input multiple-output) and beamforming - Proposed suitable Neural Network based methodologies by exploring papers on Natural Language Processing, Computer Vision, Signal Reconstruction, and Noise Cancellation - Built, tested and selected core ML models by comparing structures and results of Autoencoders, Convolutional Neural Network (CNN), and Capsule Neural Network (CapsNet) utilizing TensorFlow 1.x and Keras in Python - Enhanced core ML models to better simulate an envisioned 6G infrastructure and data transmitting process - Published the ML research and experimental results in an internal report for Huawei 6G research

- Tutored over 500 undergraduates in the course of Computer Science Software Design through office hours, tutorial sessions and final review lectures; Scored highly in the end-of-course student feedback survey - Automated assignment grading processes by developing Python scripts in Linux

Graduated with Dean's Honors