| | | 20,000+ Fresh Resumes Monthly | |
|
|
| |
Click here or scroll down to respond to this candidate Candidate's Name
Email: EMAIL AVAILABLE | Phone: PHONE NUMBER AVAILABLE
GitHub: github.com/wzhlifelover | LinkedIn: LINKEDIN LINK AVAILABLE
EDUCATION
CORNELL UNIVERSITY, New York, NY May Street Address
Master of Engineering, Major in Computer Science, GPA: 3.Street Address
Relevant Coursework: Applied Machine Learning, Cryptography, Algorithms for Application, Computer Vision
THE COOPER UNION FOR THE ADVANCEMENT OF SCIENCE AND ART, New York, NY May 2021
Bachelor of Engineering, Major in Electrical Engineering, Minor in Computer Science, GPA: 3.70
SKILLS
Programming Languages: JavaScript/TypeScript, Python, Java, Bash, C/C++, SQL | Framework: React, Node.js, PyTorch
Cloud: GCP, AWS | Database: PostgreSQL, Redis | LLM: GPT-4, LLAMA2 | Spoken Languages: Mandarin, Japanese
PROFESSIONAL EXPERIENCES
MD.AI, New York, United States Summer 2021-Now
Software Engineer (JavaScript/TypeScript, Python)
Pioneered a medical AI platform that incorporates an image viewer used for assembling training datasets and running
inference with MD.ai-developed models, along with a ChatGPT-like application for LLM-assisted clinical reporting
Led the training and deployment of machine learning models for medical image segmentation and clinical report
generation, streamlining dataset creation, annotation, preprocessing, and fine-tuning to elevate model performance
Created and maintained React components for a cross-platform clinical reporting app with AI-assisted dictation and
synchronous editing, and a medical image viewer equipped with algorithm-based automatic annotator
Developed internal GraphQL APIs for database management and external router APIs for user project, dataset, and
model I/O and progress retrieval; built Python libraries and CLI as wrappers for script-based user command execution
Engineered a knowledge injection workflow with vector databases and LangChain for GPT-4 to incorporate the latest
medical research and news, aimed at minimizing hallucinations and maximizing content relevance
Engaged in prompt engineering for LLMs to improve medical report proofreading, de-identification, and conversion of
exam findings into templated reports, augmenting reporting efficiency and compliance
Aligned with medical standards DICOM and HL7 by facilitating bidirectional transformation of regular images and texts
into specialized formats for clinical exams and reports, enabling data exchange within standardized FHIR workflows
SHANGHAI AITROX TECHNOLOGY, Shanghai, China Summer 2019
Software Development Intern (C++)
Worked on a Qt-based floating window tool for a medical reporting system, which extracts patient ID from medical
reports with OCR pipeline, enabling automatic loading of exam and patient info
Enhanced the recognition system accuracy and reliability by managing dataset synthesis, tuning the OCR model, and
applying postprocessing techniques to model outputs, thereby reducing false positive rate
Optimized UI/UX by adjusting component attributes such as app layout and colors with Qt Tools
PROJECT WORK & RESEARCH
TRAVELUP, Cornell University Spring 2022
Co-designed an internal project organization platform with Travelers Insurance using React and Python
Formulated a data schema to quantify employee skills and interests in relation to team and project requirements, then
implemented optimization algorithms to maximize member-to-team matching scores
COMPUTER VISION-BASED AI ROBOTIC ARM, The Cooper Union Spring 2021
Designed and assembled a robotic arm that can recognize and stack Jenga blocks with a success rate of 90%
Programmed and calibrated a stereo camera system in C and Python, enabling depth sensing for accurate positioning
of the robot arm tip and Jenga blocks within a 1cm error margin and integrating APIs for precise robot motor control
Prepared and annotated an image dataset of Jenga blocks and trained an object segmentation algorithm with it,
allowing the system to identify and lock onto all Jenga blocks instances and thereby create a queue for tower stacking
|
