Haotian Wang Email & Phone Number

Salt Lake City, Ut, Us

Salt Lake City, Ut, Us

Castell, an Intermountain's population health platform company.- Designing, customizing, and fine-tuning machine learning-based and rule-based predictive models for electronic medical record (EMR) big data and healthcare big data.- Healthcare actuarial cost model, trend model for healthcare trend analysis.- Generative AI - Large Language Models (LLMs) for automated transcript question-answering.- Speech recognition for converting telecom calls to documents.- Computer vision-based clinical document conversion and LLM-enabled documentation review.- Decision support.

- Designed a semantic segmentation neural network to recognize microstructures in Transmission Electron Microscopy for the fuel cladding chemical interaction (FCCI) in Python, OpenCV, and TensorFlow.- Partnered with the scientist team to optimize algorithms, conduct experiments and publish research as the first author, under the supervision of Dr. Tiankai Yao

- Developed an unsupervised clustering-based algorithm with quantitative APT data analysis in Matlab to measure the stability of structural material in nuclear reactors in real-time, which increased accuracy and decreased costs.- Evaluated and analyzed cluster number and size to determine the material life cycle.

PhD Thesis: Deep Multi-task Learning for Histopathology Image Processing, University of Idaho, under the supervision of Prof. Min Xian- Developed deep learning solutions to accelerate digital pathology (DP) clinical diagnostics and cancer research.- Successfully increased detection accuracy and generation performance by 7-11% compared to the current best models.- Designed and implemented deep CNN-based segmentation models to extract different nuclei, cells, and tissues in histopathology images using Python, OpenCV, Scikit-Image, TensorFlow, Matlab, and Linux shell.- Designed generative AI models including Generative Adversarial Networks (GANs), Variational autoencoders (VAEs), and Neural Style Transfer models for semantic and instance-level histopathology image generation using Python and Pytorch. · Unsupervised Autoencoder for Intelligent Intrusion Detection System · AI-Driven Crop-Field Monitoring · Human Health Monitoring with Motion Intelligence· Teaching Assistant for Deep Learning, Python for Machine Learning.

Atlanta, Ga

- Created a deep learning-based object detection model for capturing and detecting oil leak faults on thousands of offshore drilling platforms with a camera on an unmanned aerial vehicle using Python, TensorFlow, and OpenCV.- Partnered with the engineering team to design algorithms and develop an AI demo that achieved a 91% precision rate.

Idaho Falls

- Built and optimized a supervised linear regression-based model in R for simulating and predicting the catalytic process within the first 3 steps; achieved 80+% accuracy, reduced manual experimentation, and reduced low-quality results.

- Polycom, acquired by HP. - Mined high-quality human face data from Youtube, aggregated, organized, and pruned of the dataset with millions of image samples for face recognition, and developed detection algorithms for video conference systems using C++ and OpenCV.

Doctor Of Philosophy - Phd, Computer Science (Machine Learning)

Department of Computer Science, University of Idaho

Bachelor Of Science - Bs, Electrical Engineering