👋 Hi there! I'm Chien.
I'm an engineer at the intersection of machine learning, infrastructure, and full-stack development. At Aignostics, a medical AI startup, I build scalable ML pipelines and developer tooling that power drug discovery research. My background spans LLM fine-tuning at NIO, distributed database research at IBM, and computer vision work published at CVPR and ICML — always focused on real-world, measurable impact. When I'm not coding, I enjoy cooking and playing badminton.
Hi there! I've been trained on Chien's professional background and documentation.
Feel free to ask me any questions about Chien!
Experience
- JULY 2024 - CURRENT
Machine Learning Engineer - MLOps @ Aignostics
I work in a cross-functional team to build AI solutions for drug discovery, with a focus on scalable ML infrastructure, pipeline optimization, and internal developer tooling on Google Cloud.
- Deployed scalable ML training pipelines using Google Kubernetes Engine (GKE), reducing model training time by 10x through improvements in infrastructure and TensorFlow/PyTorch algorithms.
- Optimized Docker images and CI/CD build processes, reducing build time by 90% by refining image layers and caching Python dependencies.
- Developed an internal tool for cost-efficient GPU access on Google Cloud Platform (GCP), saving the team over 18 hours per day in development time.
- DECEMBER 2023 - MAY 2024
Software Engineer Intern - AWS @ NIO
I contributed to the NOMI team by building an end-to-end Automatic Speech Recognition (ASR) system — from data collection and model fine-tuning to an automated benchmarking pipeline and a full-stack dashboard for tracking model performance.
- Improved German ASR accuracy by 35% on internal benchmarks by fine-tuning OpenAI Whisper models and applying custom audio preprocessing in Python.
- Built a benchmarking pipeline using Apache Airflow to automate weekly ASR model evaluations and developed a front-end for data visualization, saving over 10 hours of manual work per week.
- OCTOBER 2022 - MAY 2023
Software Engineer Intern @ IBM
Worked with the cloud database team to research anomaly detection for distributed SQL databases. Completed my Master's thesis with IBM, focusing on AI-driven event detection for PostgreSQL systems.
- Designed and implemented a data-driven error detection algorithm for PostgreSQL running on Kubernetes clusters using Python and machine learning techniques.
- APRIL 2021 - SEPTEMBER 2021
Data Scientist Intern @ Teraki
Teraki provides edge AI software for vehicles and IoT devices. I developed object detection models for time-series radar data to identify vehicles and pedestrians.
- Accelerated radar data preprocessing pipelines by 10x by integrating Cython into the Python codebase.
- Developed end-to-end deep learning solutions for radar object detection using PyTorch and deployed models on edge devices.
- OCTOBER 2018 - JUNE 2020
Research Assistant @ ElsaLab
Led machine learning projects in image processing and reinforcement learning, from research to deployment. My work was published in top conferences such as CVPR and ICML.
- Developed state-of-the-art visual odometry solutions on the KITTI benchmark using TensorFlow.
- Designed distributed training strategies for ML models using on-premises servers and Google Cloud Platform (GCP).
- Implemented robot control systems for self-driving cars using C++ and Robot Operating System (ROS).
- MAY 2019 - AUGUST 2019
Student Research Assistant @ National Tsing Hua University
Led an industrial-academic collaboration project to develop an intelligent seeding machine for Malabar Chestnut.
- Designed image classification algorithms using OpenCV and Python to improve seed orientation detection and increase production rate by 90%.
- Implemented automation control systems for motors and pneumatic devices using C language and embedded systems.
Projects
Full-stack AI application that answers questions about Chien Liu using a retrieval-augmented generation (RAG) system. Searches a SQL database and vector store for relevant context, then uses an LLM to generate grounded responses. Deployed on Cloudflare Workers.
A python package that generates cartoon faces using DCGAN (Deep Convolutional Generative Adversarial Network). The package is uploaded to PyPI for easy installation. The training process is logged and visualized with Weight and Biases (W&B).
A navigation system combined with visual perception, localization, navigation, and obstacle avoidance using merely one single RGB camera.
Online vocab exercise for German A1 learner. The web app is built with plain HTML, CSS, and JavaScript.
Image classification using VGG19 to detect status of seeds, combining with self-developed machine, improving production rate by 90%.
Full-stack AI application that answers questions about Chien Liu using a retrieval-augmented generation (RAG) system. Searches a SQL database and vector store for relevant context, then uses an LLM to generate grounded responses. Deployed on Cloudflare Workers.
A python package that generates cartoon faces using DCGAN (Deep Convolutional Generative Adversarial Network). The package is uploaded to PyPI for easy installation. The training process is logged and visualized with Weight and Biases (W&B).
A navigation system combined with visual perception, localization, navigation, and obstacle avoidance using merely one single RGB camera.
Online vocab exercise for German A1 learner. The web app is built with plain HTML, CSS, and JavaScript.
Image classification using VGG19 to detect status of seeds, combining with self-developed machine, improving production rate by 90%.