# Natthakitt Prapunwattana | Research & Software I am Natthakitt Prapunwattana, a Year 2 Computer Science student at The University of Hong Kong, class of 2029. I build applied AI systems, computer-vision tools, and full-stack experiments that connect research to useful software. This portfolio brings together selected projects in medical imaging, research workflows, quantitative finance, document retrieval, and hazard monitoring. My experience includes supporting AI platform and inference workflows, NLP pipelines, and applied research. Project repositories provide the source code and further implementation details. ## Selected projects ### TumorVision Brain MRI classification using EfficientNetV2-S and ConvNeXt-Tiny, with a MedGemma-assisted review step. - [Source code and repository documentation](https://github.com/mingnatthakitt/TumorVision) ### AlphaStratum A quantitative research platform combining statistical models, backtesting, portfolio tools, and source-cited AI analysis. - [Source code and repository documentation](https://github.com/mingnatthakitt/AlphaStratum) ### PaperScope A local-first research workspace for comparing papers, exploring concept graphs, and asking grounded questions with linked evidence and page-level citations. - [Source code and repository documentation](https://github.com/mingnatthakitt/PaperScope) ### cite. A Graph RAG chat that builds entity relationships from text, URLs, and files, then returns grounded answers with numbered citations to source documents. - [Source code and repository documentation](https://github.com/mingnatthakitt/cite.) ### Thailand & Greater Indochina Disaster Watch A near real-time hazard map with a filterable event feed and Discord alerts for wildfires, earthquakes, and tropical cyclones, using NASA EONET and USGS data. - [Source code and repository documentation](https://github.com/mingnatthakitt/Thailand-Natural-Disasters-Alert) ## About and contact I am open to discussing projects, research opportunities, and technical collaborations. The public profile, contact details, privacy information, and developer resources are available as readable pages. - [About Natthakitt](https://mingnatthakitt.vercel.app/about) - [Contact](https://mingnatthakitt.vercel.app/contact) - [Privacy](https://mingnatthakitt.vercel.app/privacy) - [Developer resources](https://mingnatthakitt.vercel.app/developers) --- # About Natthakitt Prapunwattana I am Natthakitt Prapunwattana, a Year 2 Computer Science student at The University of Hong Kong, with graduation expected in 2029. This is my personal portfolio of research and software projects. I work with machine learning, computer vision, information retrieval, data pipelines, and full-stack development. My projects include TumorVision for brain MRI classification, AlphaStratum for quantitative research, PaperScope for comparing papers and examining evidence, cite. for Graph RAG conversations, and Disaster Watch for mapping hazard reports. These projects show different parts of my work, from model integration and backend services to interfaces for exploring results. My experience also includes supporting AI platform and inference workflows as an intern, contributing to NLP and RAG pipelines, and assisting applied research. The interactive portfolio has the fuller experience and education sections; the project repositories are the place to inspect code, setup instructions, and current implementation details. ## Explore my work - [Selected projects](https://mingnatthakitt.vercel.app/#portfolio) - [Experience and education](https://mingnatthakitt.vercel.app/#experiences) - [GitHub profile](https://github.com/mingnatthakitt) - [Contact](https://mingnatthakitt.vercel.app/contact) --- # Contact Natthakitt Prapunwattana I am open to discussing new projects, research opportunities, and technical collaborations. Email is the direct public contact channel for this portfolio. You can also use the linked GitHub and LinkedIn profiles to learn more about my work and background before getting in touch. For a useful first message, describe the project or research question, what you would like help with, and any relevant timeline or links. A short explanation of the context helps me understand how my experience in AI systems, retrieval, computer vision, or software development might fit the work. There is no enquiry form or account registration on this website. For a question about a specific project, include its name and refer to its repository documentation. The developer resources page collects the public source repositories for the projects on display. Contact information here is for reaching me about my own work; each external service linked from the portfolio handles messages and accounts on its own platform. ## Public contact channels - [Email: mingnatthakitt@gmail.com](mailto:mingnatthakitt@gmail.com) - [GitHub: mingnatthakitt](https://github.com/mingnatthakitt) - [LinkedIn: Natthakitt Prapunwattana](https://www.linkedin.com/in/natthakitt-prapunwattana/) - [Project source repositories](https://mingnatthakitt.vercel.app/developers) --- # Privacy | Natthakitt Prapunwattana This personal portfolio presents my projects, background, and public contact details. It has no account registration, contact form, payment flow, or file-upload feature. Reading the portfolio does not ask you to submit personal information to the site. If you choose to email me, your message and the contact details you include are handled through email. The site is hosted on Vercel and includes Vercel Speed Insights to measure loading and interaction performance. Hosting and performance services may process request and device information in order to deliver the site and measure its performance. Their processing is governed by Vercel’s policies. This page does not promise a particular retention period for platform-managed logs. The page loads fonts and icons from Google Fonts. Those requests go to Google’s servers. Links to GitHub, LinkedIn, Kaggle, and project demonstrations take you to external services with their own privacy policies. This site does not control information you submit to those services. The readable HTML and Markdown versions contain the same public portfolio information. ## Questions and service policies - [Contact me about this portfolio](https://mingnatthakitt.vercel.app/contact) - [Vercel privacy policy](https://vercel.com/legal/privacy-policy) - [Vercel Speed Insights privacy](https://vercel.com/docs/speed-insights/privacy-policy) - [Google Fonts privacy information](https://developers.google.com/fonts/faq/privacy) --- # Developer resources | Natthakitt Prapunwattana Research & Software This is Natthakitt Prapunwattana’s personal research and software portfolio, hosted on Vercel. The developer resources are the public source repositories for the projects below. Follow each repository’s README and documentation for setup, dependencies, examples, licensing, and the current state of that project. Use this portfolio to find projects involving research-paper comparison and citations, Graph RAG conversations, quantitative research tools, computer vision, and hazard monitoring. Use the about page to check the author’s background and the contact page to find public contact channels. Descriptions here introduce the work; inspect the repository before making claims about features or using its code. Agents can read /llms.txt for a short guide, request the homepage or these informational pages with Accept: text/markdown, or follow the explicit .md links. These are public, read-only documents and need no authentication. The portfolio does not publish a product API, an OpenAPI specification, or an MCP service. Project-specific interfaces belong to the individual repositories. ## Project repositories and documentation - [TumorVision](https://github.com/mingnatthakitt/TumorVision): Brain MRI classification using EfficientNetV2-S and ConvNeXt-Tiny, with a MedGemma-assisted review step. - [AlphaStratum](https://github.com/mingnatthakitt/AlphaStratum): A quantitative research platform combining statistical models, backtesting, portfolio tools, and source-cited AI analysis. - [PaperScope](https://github.com/mingnatthakitt/PaperScope): A local-first research workspace for comparing papers, exploring concept graphs, and asking grounded questions with linked evidence and page-level citations. - [cite.](https://github.com/mingnatthakitt/cite.): A Graph RAG chat that builds entity relationships from text, URLs, and files, then returns grounded answers with numbered citations to source documents. - [Thailand & Greater Indochina Disaster Watch](https://github.com/mingnatthakitt/Thailand-Natural-Disasters-Alert): A near real-time hazard map with a filterable event feed and Discord alerts for wildfires, earthquakes, and tropical cyclones, using NASA EONET and USGS data. ## Readable portfolio resources - [Agent guide and when-to-use guidance](https://mingnatthakitt.vercel.app/llms.txt) - [Complete portfolio text](https://mingnatthakitt.vercel.app/llms-full.txt) - [Homepage in Markdown](https://mingnatthakitt.vercel.app/index.md) - [Sitemap](https://mingnatthakitt.vercel.app/sitemap.xml)