Consulting 

Mathimatikos.xyz (January 2025 - Dec 2025) Conceived, designed, and launched mathimatikos.xyz, an AI-powered web application that helps university students construct clear, step by-step solutions to complex problems in mathematics, statistics, finance, physics, and computer science. Designed and implemented the full solution lifecycle—from problem definition and user needs analysis to model logic, deployment, and iteration—despite no formal software engineering background. Developed AI-assisted solution generation workflows that guide users through structured mathematical reasoning rather than producing opaque answers, emphasizing transparency and learning outcomes. Integrated responsible AI principles by prioritizing explainability, step validation, and controlled output generation to support ethical AI use in education. Iteratively refined the platform based on direct user feedback from international students at bachelor’s and master’s level, improving clarity, usability, and learning effectiveness. Translated advanced analytical reasoning into an accessible, user-centric digital product, demonstrating strong business translation and product thinking.

AI-driven repository designed to support disaster risk reduction (DRR) efforts. (Oct 2024 - Jun 2025) Honored with an international award for my work on AI-driven projects designed to support disaster risk reduction (DRR) efforts. My work presented at the Global Platform for Disaster Risk Reduction (GP2025) in Geneva, a multi-stakeholder forum mandated by the UN General Assembly. Young innovators were invited to propose solutions combining science, technology, and traditional knowledge to enhance early warning systems. The selected projects are now featured in a virtual exhibition and also showcased at the Global Multi Stakeholder Forum on Early Warnings for All, taking place in Geneva on 2–3 June 2025. This repository contains a collection of AI-driven projects designed to support disaster risk reduction (DRR) efforts. These projects leverage machine learning and deep learning to predict and mitigate the impact of extreme weather events such as tropical storms, droughts, and floods. By utilizing data-driven approaches, these models provide early warning systems (EWS) to aid decision-makers in disaster preparedness, mitigation, and response. Each project focuses on a specific disaster scenario, incorporating predictive modeling, feature engineering, and real-world deployment strategies.

BetterPrompt – A browser add on (to be launched on March 2026) Designed end-to-end a high-performance browser extension that acts as a real-time prompt engineer. It captures raw user intent from any webpage and transforms it into a structured, expert-level query using established frameworks (like RTF and Chain-of-Thought). Instead of one-size-fits-all, the app classifies input (Academic, Technical, Creative) and applies specialized optimization logic. Users highlight text and instantly "bridge" it to ChatGPT, Claude, or Gemini without manually switching tabs or copy-pasting. Built with a Railway-hosted proxy to keep API keys secure and a Supabase-backed usage tier for sustainable monetization.