Work experience

• AI Engineer

  Etiqa Insurance and Takaful

  2025.06-Current

  (a year)

  • Contributed to the replacement of a rule-based Straight-Through Processing (STP) workflow with an AI-driven approach to improve accuracy, reduce manual errors, and speed up claims processing. • Applied Multimodal and Vision-Language Models (MLLM/VLM) for OCR text extraction, document relevance and validity checks, document type classification, and language detection with automated translation to English. • Used Large Language Models (LLMs) to extract structured claim information from unstructured document text. • Performed data labeling on travel claim datasets (travel delay and missedconnection) to support model training and evaluation. • Analyzed model outputs and iteratively refined prompts to improve extraction accuracy and overall system performance. • Configured and deployed Vision-Language Models (VLMs) with vLLM for high-throughput inference on a UAT environment.
• Vision Engineer I

  Pentamaster Corporation Berhad (Official)

  2023.07-2025.04

  (2 years)

  (1 year 10 months) • Developed and integrated a new calibration tool for tilted camera setups in 2 months and AI Edge Learning tools (Classify, Segment & OCR) in 4 months with WPF in C# • Added a user-friendly feature for adding/removing vertices in polygon regions using k-d tree algorithm and vector dot product using C# • Leveraged 3rd party Deep Learning solutions to evaluate medical pills, ferrite core, CMOS and wafer die images, handling data labelling, training, parameter tuning, and reporting • Conducted AOI on medical pills, identifying solutions for cosmetic defect detection with rule-based, Deep Learning, AI Edge Learning, and 3D line scan technologies • Implemented parallel processing using C#, reducing the Segment tool's execution time by up to 40%
• Undergraduate Research Assistant

  Universiti Sains Malaysia

  2021.08-2021.11

  (4 months)
• Customer Service Representative

  transcosmos inc.

  2019.02-2019.08

  (7 months)

  • Delivering excellent customer service to customer via phone (inbound). • Providing accurate information on products and services to customer. • Maintaining high degree of integrity and confidentiality – always keeping customers’ information confidential. • Achievements: Solved an average of 30 calls per day and awarded the Best CSAT Award.

Educational experience

• Universiti Sains Malaysia

  (Minor)

  2019.09-2023.03

  (4 years)

Certificates