Nur Hapinder Binti Abdullah

• Design a LLM model that can identify risk from semantic search for a quicker risk identification across multiple projects. • Organize and participate ad hoc workshop for client and MODEC internal departments. • Work with the team to implement a Natural Language Processing (NLP) solution improved company chatbot accuracy to 95%. • The application was built using Java/J2EE (Spring Boot, Hibernate) and integrated with both Oracle DB and MSSQL to support complex risk analytics and reporting.

Data Analytics/ AI: • Identify and define KRI relevant to the organization’s objective and risk appetites. • Collaborate with different stakeholders in designing and implementing end-to-end AI solutions to address business challenges. • Leverage technology tools for efficient KRI monitoring and reporting. • Continuously monitoring KRI and provide regular reports to management, highlighting emerging risk and trends. • Work in a team on LLM ecosystem, including frameworks like LangChain, and vector databases.

• Used Tensorflow to train model for CNN. • Built ETL pipelines in Databricks using PySpark to process manufacturing & test data for yield analysis and defect classification in semiconductor production. • Designed and deployed machine learning pipelines to improve wafer yield & reduce defect density in back-end processes. • Collaborated with process and yield engineers to translate domain knowledge into model features, validation plans, & root-cause analyses.