A little about who I am, where I come from, and what drives me.
I'm a data scientist who works with AI and data to build practical solutions that turn information into insights and intelligent systems.
My expertise lies in machine learning, data science, and applied AI. I enjoy exploring how data can be transformed into models that support better decisions and solve real-world problems.
I'm particularly interested in looking for opportunities to engage and learn beyond AI — whether that's through new problem domains, cross-functional collaboration, or emerging technologies.
LOCATION
Malaysia
EDUCATION
UTAR - Bachelor of Information Systems (Hons) Digital Economy Technology (CGPA 3.88)
My professional journey so far — latest first.
Kyndryl Malaysia
Building intelligent agentic systems and automated enterprise pipelines. Designed a stateful multi-agent orchestrator with LangGraph & Vertex AI for autonomous booking workflows, implemented a document processing pipeline using Azure Document Intelligence for compliance validation, and developed an AI contract management system leveraging AWS Bedrock for draft generation, redlining, and legal retrieval.
EY Technology Consulting
Contributed to an AI SQL Agent leveraging NLP text-to-SQL techniques to enhance structured data extraction from databases. Supported System Integration Testing for a GenAI project and assisted in frontend development of a GenAI chatbot using React JS.
Technologies and tools I work with daily.
Projects that showcase my expertise in AI, NLP, and cloud technologies.
Designed and deployed a stateful multi-agent orchestrator using LangGraph and GCP Vertex AI, implementing conditional routing and supervisor agent patterns to automate internal bookings. Leveraged Vertex AI Search for enterprise data retrieval, managing agent memory and session states via Cloud Firestore to handle asynchronous user intents. Increased operational efficiency by streamlining booking workflows and reducing manual intervention through autonomous agent orchestration.
Developed an AI contract management system leveraging AWS Bedrock and AWS S3-based vector storage across three core modules: a draft generation agent for standardized agreements, a redlining agent for non-compliant clauses via semantic risk mapping, and an interactive legal assistant chat agent for intelligent contract retrieval and analysis. Accelerated legal review cycles and proactively identified compliance risks.
Implemented an automated document processing pipeline utilizing Azure Document Intelligence (Read & Layout models) to extract unstructured data from complex loan applications. Integrated Azure OpenAI with the Structured Outputs API to enforce strict schemas for cross-validation against compliance datasets, reducing schema validation errors and automating risk flag generation.
Created a system for a banking application to provide real-time, context-aware resolutions to user queries. Leveraged Retrieval-Augmented Generation combined with hybrid searching using BM25 and vector search powered by Azure OpenAI. Improved query resolution accuracy by grounding the LLM in specific banking documentation through semantic search and vector indexing.
Enhanced structured data extraction by allowing users to query databases using natural language. Developed an AI SQL Agent using NLP text-to-SQL techniques with LLM to understand human language and convert it to SQL queries for retrieving relevant data from Postgres. Implemented few-shot prompting to improve accuracy and performance.
Developed a machine learning classification model to automate credit score prediction. Applied Random Forest and Logistic Regression algorithms with comprehensive model evaluation using AUC-ROC and F1 Score metrics. Implemented using Python, Scikit-Learn, Matplotlib, and Power BI for risk assessment and visualization.
Industry-recognized credentials across major cloud platforms.
4 certifications
2 certifications
1 certification
I'm always open to new opportunities and conversations about AI and data science. Feel free to reach out through any of the channels below.