Tianchang Liu

Led the optimization and deployment of computer vision models for touchless checkout, achieving 98.5% SKU recognition accuracy. Developed a synthetic data pipeline and implemented MLOps practices, enhancing model deployment efficiency and reducing latency. Created generative AI workflows and engineered data pipelines for healthcare compliance.

Designed decentralized ML infrastructure protocols and frameworks, optimizing reinforcement learning across thousands of GPUs. Developed LLM and NLP advancements, creating pipelines for research papers and implementing smart contracts for cost efficiency.

Improved autonomous vehicle perception through advanced model training and optimization. Built data pipelines for sensor data processing and orchestrated automated retraining workflows. Deployed models to fleets with low inference times and developed simulation environments.

Developed machine learning solutions across various domains including robotics and healthcare AI. Built ROS nodes for robotic path planning and deployed models for medical data analysis. Scaled backend systems for SaaS applications.

Developed machine learning models for product recommendations and analyzed customer data to improve accuracy. Built ETL pipelines for processing large datasets efficiently.