I'm a Computer Science student based in Johor Bahru, Malaysia, passionate about building AI-powered solutions and turning data into actionable insights. From enterprise AI agents to real-time scam detection systems, I love tackling real-world problems under pressure β especially in hackathons.
π₯ KAWAL App β 1st Place
APC Dewan Belia Perak
A civic engagement platform designed to increase youth participation in state-level governance and community reporting.
TypeScript React CSS
Lovable x AWS Hackathon β Track 2: Clinician Copilot
An AI-powered multi-role health triage platform that connects patients, medical officers, and pharmacists in one seamless clinical workflow.
- π€ AI consultation via Google Gemini 2.5 Flash (OpenRouter) β dynamic symptom interviews with ESI triage scoring (levels 1β5) and real-time emergency detection
- πΊοΈ Interactive clinic finder with live queue management (Leaflet map)
- π Bilingual β Bahasa Malaysia & English AI consultations
- π Deployed on Cloudflare Pages with Supabase backend
React 19 TypeScript TanStack Router Vite Supabase Cloudflare Pages Gemini 2.5 Flash
Microsoft Agents League Hack @ AI Skills Fest
An enterprise AI agent for Microsoft 365 Copilot serving Risk Analysts and Financial Controllers with governed, AI-driven portfolio insights.
- π οΈ 7 MCP tools β real-time portfolio KPIs (ChurnRate, NPLRatio, LossRecoveryRate), churn driver analysis, at-risk account prioritisation
- π Microsoft Fabric IQ semantic model integration via DAX queries
- π Governed write operations secured by Microsoft Entra ID OAuth 2.1 (RS256 JWT)
- π΄ Adaptive Card v1.6 widgets rendered natively in Copilot & Teams
- π³ Fully containerised with Docker
Python FastMCP Microsoft Fabric IQ Microsoft Entra ID Docker Adaptive Cards
Codex Community Hackathon Kuala Lumpur 2026
A real-time scam protection system targeting vulnerable communities β parents, grandparents, and anyone less tech-savvy β built on six specialized AI agents.
| Agent | Function |
|---|---|
| π Call & Voice Note Scanner | Real-time scanning of incoming calls and audio messages |
| ποΈ Cloned Voice Detector | Identifies AI-generated voice impersonation |
| πΌοΈ Deepfake Image Detector | Flags manipulated images used in scam attempts |
| π¬ Message Pattern Analyser | Detects suspicious language patterns in text |
| π¦ Verdict Engine | Issues clear Safe / Caution / Danger verdicts with explanations |
| πΊοΈ Live Heatmap | Real-time scam activity map across Malaysia |
π Multilingual β Bahasa Malaysia, English, Mandarin Chinese & Tamil
Python Multi-Agent AI NLP Computer Vision Deepfake Detection
End-to-end ML pipeline to classify fraudulent transactions with high precision and recall.
- π° Mitigated approximately $46,758 in potential financial losses
- π Built an interactive Power BI dashboard with transaction-level scatter plots and risk distribution charts
Python Scikit-learn XGBoost Pandas Power BI (DAX)
Modular churn analytics tool to identify high-risk customer segments and drive retention.
- πΊοΈ Geographic dashboards tracking churn rates by region for localised interventions
- π Feature importance analysis to pinpoint top drivers of customer attrition
Python Scikit-learn Pandas Tableau Kaggle
| Category | Tools & Technologies |
|---|---|
| Languages | Python, TypeScript, JavaScript, SQL, Java, C++, PHP, HTML, CSS |
| Frameworks & Libraries | React 19, TanStack Router, Vite, Pandas, NumPy, Scikit-learn, XGBoost, Leaflet |
| AI & ML | Multi-Agent AI Systems, NLP, Machine Learning Pipelines, Gemini 2.5 Flash, Deepfake & Voice Detection, FastMCP |
| Data & Visualization | Power BI (DAX), Tableau, Microsoft Fabric IQ, Geographic Dashboards, Adaptive Cards |
| Cloud & DevOps | Supabase, Cloudflare Pages, Docker, Microsoft Azure (Entra ID), Git / GitHub |
- β Google Data Analytics Professional Certificate β Google
- β³ IBM AI Engineering Professional Certificate (In Progress) β IBM
IT Management Intern Β· KPJ Johor Specialist Hospital (Sep 2024 β Mar 2025)
- Developed a Hospital Computer Mapping Website β centralized inventory system to track and map all IT assets hospital-wide
- Resolved hardware/software issues across departments; configured LAN network connectivity
- Produced corporate-level video content for internal hospital communications
