Software Engineer

Wan-Chu (Winnie) Lin

631-***-**** ******.******@*****.*** linkedin.com/wan-chu-lin

Education

Stony Brook University Expected Graduation: Dec 2025 M.S. in Computer Science New York, US

• Relevant Coursework: System Security, Analysis of Algorithms, Data Science Fundamentals National Taiwan University Sep 2017 - Jun 2019

M.S. in Biostatistics (GPA: 3.95) Taipei, Taiwan

National Taiwan University Sep 2012 - Jun 2016

B.S. in Public Health Taipei, Taiwan

Experience

E.SUN Bank Mar 2022 – Jul 2024

Software Engineer Taipei, Taiwan

• Developed Kubernetes-based AI web services for 8,000 employees (OCR, TTS, chatbots, meeting services). Led lifecycle management, boosting operational efficiency by 30% and reducing paperwork by 50%.

• Increased OCR efficiency by 15% through the utilization of custom templates, thereby accelerating the processing of over 10,000 daily documents and diminishing the need for manual data entry by 4 hours per week.

• Designed distributed backend with Redis caching, Kafka processing, and gRPC communication, reducing response time by 60%, increasing throughput by 40%, and handling 3x more users.

• Developed the ”Easy-to-SQL” Python package leveraging os.Pipe for efficient database data transfer and cchieved 50% lower memory usage and 33% faster processing, boosting EDA efficiency and Airflow DAG performance.

• Developed an advanced dispatch system on GCP utilizing Pub/Sub, capable of processing 700,000 tasks per hour. Employed Terraform to ensure uniform infrastructure deployment across various environments. Cohesion Data May 2020 – Feb 2022

Software Engineer Taipei, Taiwan

• Implemented an ETL pipeline to handle 200,000+ daily DICOM files from S3 to MongoDB with parallel processing, reducing processing time from 6 hours to an hour and enabling real-time data analysis for government personnel.

• Developed an interactive dashboard using React, Cube.js, and MongoDB to detect fraudulent medical images, reducing staff review time by 80% and cutting costs by avoiding erroneous payments.

• Developed a nested QA model, leveraging BERT embeddings, to extract critical information such as disease, location, and attributes from electronic health records. Achieved an F1-score of 0.867. Gokube Sep 2019 – Mar 2020

Data scientist Taipei, Taiwan

• Developed a GCP-based data pipeline with BigQuery, DataPrep, and Data Studio for daily updates of prediction models, improving strategic planning insights by 30%.

• Trained a statistical model to predict hourly demand and allocate bikes, leading to a 12% daily trip increase. Projects

AI-assisted Thoracic X-ray Imaging for Rural Medical Outreach Vans Vue.js, Fastapi, PyTorch, Docker

• Collaborated with physicians Hualien Tzu-Chi Hospital to train a lung segmentation model and a thoracic anomaly multi-label classification model.

• Integrated AI algorithms with the hospital’s HIS system, streamlining the identification of critical abnormalities in imaging.

Skills

Programming Languages: C++, Python, TypeScript, R

Data & ML: Airflow, PostgreSQL, MongoDB, Redis, Elasticsearch, TensorFlow, PyTorch Frameworks: React.js, Vue.js, Express.js, Node.js, Fastapi DevOps & Systems: Git, Docker, Prometheus, Grafana, Ansible, Terraform, Kubernetes, Unix/Linux

Contact this candidate