o
ooi_zheng_yang

Yang

@ooi_zheng_yang

Data Scientist, Software Engineer

Malesia
Inglese, Cinese, Malese
Alcune informazioni sono riportate in lingua inglese.
Chi sono
I am a Data Scientist with a strong foundation in software engineering, specializing in ML and LLM systems within the Insurtech sector. Developed and deployed production-grade ML systems, including computer vision classifiers and document extraction solutions leveraging OCR and self-hosted LLM models. My portfolio includes AI-powered side projects that incorporate RAG, LLM and VLM. Hit me up if you have any ideas or solutions you want to bring it to reality!... Continua a leggere

Competenze

o
ooi_zheng_yang
Yang
offline • 

Consulta i miei servizi

Siti Web e software IA
I will create or integrate ai to websites

Esperienza lavorativa

Silverlake_Axis

Junior Data Scientist & Solutions

Silverlake Axis • Full time

Jun 2026 - Present • 5 mos

1. Built a document classifier trained with 25k images in a week, and achieved 99% accuracy on unseen data and deployed it as a microservice 2. Engineered an OCR + LLM-powered document extraction pipeline supporting multiple document types and languages 3. Evaluated and benchmarked LLMs and VLMs across diverse tasks, assessing output quality, accuracy, and consistency to identify the most suitable models for different use cases. 4. Experienced in improving damage detection and cost prediction models for two core company products through model retraining, improved data quality, and feature engineering, achieving 5% improvement in both model performance. 5. Delivered product presentations and technical demonstrations to non-technical clients 6. Gather and clarify requirements with non-technical clients for their required rules and workflows for their systems

Venture

Software Engineer

Venture • Part time

Nov 2024 - May 2025 • 6 mos

1. Built a full-stack system supporting end-to-end order and manufacturing across 6 workflows stages 2. Engineered ETL pipeline for data processing, transformation, and validation, reduced ~20% inconsistent data 3. Developed automation tools to support data ingestion and ETL workflows to eliminate manual errors 4. Implemented backend APIs, workflow validation logic, and data constraints to ensure system correctness 5. Collaborated with engineers to debug, test, and improve system stability in production environments