


Hello, I am
I study Computer Science, majoring in AI, at Multimedia University -- Dean's List every semester, CGPA 3.93.

I specialize in Artificial Intelligence, data engineering and full-stack development. I build intelligent, agentic systems that automate complex workflows and bridge the gap between heavy technical architecture and real-world business impact.
1 / 2Where it started. Software engineering foundations, database architecture and the first real web applications.

A full-stack storefront for fitness gear: Google OAuth, multi-method checkout and role-based admin dashboards.

An automated customer-feedback pipeline on Apps Script and OpenAI โ sentiment analysis, Gmail follow-ups, Notion sync.

Smart-contract lease agreements with escrowed payments and ZK privacy across MasChain, Scroll and Aptos.
1 / 2First industry role. Enterprise software workflows, code-architecture improvements and agile delivery.

An on-chain credit-lending standard using World ID and TEE-based MPC wallets for under-collateralized loans.

Majoring in Artificial Intelligence. Machine learning, computer vision, NLP and expert systems.

Competitive programming, algorithmic design and problem solving under time pressure.

A Telegram prediction market built on ElizaOS with on-chain token escrow settlement.
1 / 2A chaotic financial agent: multi-modal OCR on receipts plus Agentic RAG for passive-aggressive spending roasts.

A Flutter app using the Vision API to scan barcodes, track expiry, suggest AI recipes and gamify food donations.
1 / 2An end-to-end judging platform with a multi-LLM auto-review pipeline and OCR preprocessing.

An adaptive traffic-light system replacing fixed timers with real-time flow optimisation via YOLOv8.

A privacy-preserving Web3 platform on Oasis Sapphire matching novice traders with experienced professionals.

A business analytics platform with AI-driven workflows and 12 custom tools wired through MCP.
1 / 2Immersive language learning combining real-time object detection with speech pronunciation scoring.
1 / 2A dual-pathway RAG framework over LLMs to make credit-risk assessment more accurate and transparent.
1 / 2An enterprise recruitment ecosystem using Gemini 2.0 Flash for PDF parsing and WebGL 3D avatars for bias-free interviews.
1 / 2A decoupled threat-detection system: a Puppeteer sandbox for URL detonation plus a multi-agent Gemini pipeline.

Dual-RAG and hybrid neuro-symbolic architecture turning LLMs into verifiable, policy-anchored decision engines.
1 / 2Eliminating AI hallucination in loan advisory and cutting inference cost by 90% with policy-anchored retrieval.

A multi-agent business planning platform scoring ideas with expert AI agents and a 72-hour outcome simulation.
1 / 2Co-built ShogunOS, a multi-tenant AI operations platform running isolated departmental agents per company โ architecture through to live customer deployment.
Major in Artificial Intelligence


I specialize in translating complex requirements into responsive web and mobile applications using TypeScript and Next.js. Through competitive hackathons, I have mastered the art of building data-rich dashboards and cross-platform mobile apps with Flutter, focusing on Core Web Vitals, strict type safety and seamless UI/UX design in Figma. My approach emphasizes reusable component architecture and "Local-First" data strategies to ensure performance even in low-connectivity environments.
Move across the grid โ every cell is a tool from the projects below.
Developed a multi-agent business planning platform that scores ideas based on user inputs, leverages expert AI agents for strategic feedback and CEO-level decisions and features a 72-hour visual simulation to forecast business outcomes.
Developed an Intelligent Loan Advisory & Compliance Matching Platform utilizing a Dual-RAG and Hybrid Neuro-Symbolic architecture to eliminate AI hallucinations , transforming LLMs into verifiable, policy-anchored decision engines that reduce inference costs by 90%.
Developed an Intelligent Loan Advisory & Compliance Matching Platform utilizing a Dual-RAG and Hybrid Neuro-Symbolic architecture to eliminate AI hallucinations , transforming LLMs into verifiable, policy-anchored decision engines that reduce inference costs by 90%.
Developed an automated business refinement platform that de-risks entrepreneurship for beginners by using multi-cultural agent personas and adversarial gap analysis to generate actionable 48-hour launch blueprints and 10-year vision maps.
Architected a decoupled threat detection system that uses a Node.js/Puppeteer sandbox for URL detonation alongside a multi-agent Gemini 2.5 pipeline to automatically execute network and financial defense protocols.
Developed an end-to-end enterprise recruitment ecosystem utilizing Gemini 2.0 Flash for PDF parsing and WebGL for 3D AI Avatars to conduct bias-free interviews.
Developed a dual-pathway Retrieval-Augmented Generation (RAG) framework using Large Language Models to enhance the accuracy and transparency of credit risk assessments for financial institutions.
Created an immersive language learning platform combining real-time object detection (AWS Rekognition) and advanced speech pronunciation assessment.
Built a scalable business analytics platform with AI-driven workflows, automated reporting and 12 custom tools integrated via the Model Context Protocol (MCP).
Engineered a privacy-preserving Web3 platform using Oasis Sapphire and Gemini AI to securely match novice traders with experienced professionals.
Pitched an AI-driven adaptive traffic light system that utilizes YOLOv8 computer vision and machine learning to replace fixed-timer signals with real-time, dynamic flow optimization.
Developed a Flutter mobile app utilizing the Vision API for barcode scanning to track food expiry dates, suggest AI recipes and gamify surplus food donations.
Built an end-to-end judging platform featuring a multi-LLM (OpenAI, Gemini, Claude) auto-review pipeline and OCR preprocessing for evaluating competition proposals.
Developed a chaotic financial agent that uses multi-modal vision to OCR receipts and an Agentic RAG framework to deliver "Passive-Aggressive" roasts or "Hype Man" praise based on spending habits.
Designed a Telegram-based prediction market with ElizaOS and on-chain token escrow settlement for novice and professional traders.
Demonstrated exceptional competitive programming, algorithmic design and problem-solving abilities among hundreds of participants.
Recognized for building an on-chain credit lending standard that utilizes World ID and TEE-based MPC wallets to enable under-collateralized loans across multiple chains.
Built an automated customer-feedback pipeline using Google Apps Script and OpenAI to perform real-time sentiment analysis, Gmail follow-ups and Notion data synchronization.
Developed a Web3 rental marketplace featuring smart-contract lease agreements, escrowed payments and ZK-privacy across MasChain, Scroll and Aptos.
Contributed to enterprise software development workflows, integrating system improvements, optimizing code architecture and collaborating in an agile tech environment.
Developed a full-stack e-commerce platform for fitness gear featuring Google OAuth, multi-method checkouts (PayPal/QR) and role-based admin dashboards.















Get a complete overview of my technical skill, academic background and hackathon achievements.
Download PDF