<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Projects on Daniel Hori Davila</title><link>https://danhdav.github.io/projects/</link><description>Recent content in Projects on Daniel Hori Davila</description><generator>Hugo</generator><language>en</language><lastBuildDate>Sat, 11 Apr 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://danhdav.github.io/projects/index.xml" rel="self" type="application/rss+xml"/><item><title>Disaster Assessment</title><link>https://danhdav.github.io/projects/disaster-vlm/</link><pubDate>Sat, 11 Apr 2026 00:00:00 +0000</pubDate><guid>https://danhdav.github.io/projects/disaster-vlm/</guid><description>&lt;p>This was my &lt;a href="https://github.com/danhdav/Disaster-Detection-VLM">senior capstone group project&lt;/a>, and a crucial opportunity for me to dive into Langchain and OpenRouter more.&lt;/p>
&lt;p>It consists of a dashboard map showing satellite before-and-after disaster images and their respective locations. We bench-marked and integrated a Vision Language Model to predict and show the severity of each disaster.&lt;/p>
&lt;p>The dashboard contains an integrated chat bot using Retrieval Augmented Generation. Yes, I understand there are now better methods to fetch data now instead of text chunking, but this was more about prioritizing the MVPs and deployment.&lt;/p></description></item><item><title>Taxmaxx</title><link>https://danhdav.github.io/projects/taxmaxx/</link><pubDate>Mon, 09 Mar 2026 00:00:00 +0000</pubDate><guid>https://danhdav.github.io/projects/taxmaxx/</guid><description>&lt;p>&lt;a href="https://www.hackai.org/">HackAI&lt;/a> was my first hackathon where I paired up with a random group of people. In honor of tax season (RIP), we decided to make a &lt;a href="https://devpost.com/software/project-v9c8wl">personal tax assistant&lt;/a>. The app was tailored towards inexperienced filers who want a quick and private way to optimize and analyze their tax metrics by simply uploading their tax forms.&lt;/p>
&lt;p>We used the &lt;a href="https://learn.microsoft.com/en-us/azure/ai-services/document-intelligence/prebuilt/tax-document?view=doc-intel-4.0.0">Azure Document Inference API&lt;/a> for this POC, which supports all 800 tax form types from their extensive fine-tuned models. I made a &lt;a href="https://github.com/danhdav/HackAI-2026/tree/main/backend/parser">script&lt;/a> that implemented PPI (personal identifiable information) redaction locally.&lt;/p></description></item><item><title>Tool Crib</title><link>https://danhdav.github.io/projects/tool-crib/</link><pubDate>Thu, 11 May 2023 00:00:00 +0000</pubDate><guid>https://danhdav.github.io/projects/tool-crib/</guid><description>&lt;p>I took an engineering elective that caters towards community service projects (you can find more information here). Normally you would expect to be grouped with upperclassmen in a project as a first-year, but random selection put me in a team of individuals with little to no experience as well, so we all had to shoulder a lot more responsibility.&lt;/p>
&lt;p>The fabrication shop on campus that is part of UTDesign needed an administrative tool-tracking app for their hardware tools. People would check out these tools via a paper log and sometimes stole from them. I proposed a MERN stack (and used MySQL instead of MongoDB), though later on I realized that we could have easily done something more lightweight like LAMP. It is what it is though. At least we can proudly say that we were the &lt;a href="https://sites.utdallas.edu/epics-utdesign-tool-crib/">founding software developers&lt;/a> of this project, as the development of it &lt;a href="https://github.com/UTDallasEPICS/UTDesign-Tool-Crib">extended&lt;/a> into later semesters for people to continue optimizing.&lt;/p></description></item><item><title>Summer</title><link>https://danhdav.github.io/projects/summer/</link><pubDate>Fri, 28 Apr 2023 00:00:00 +0000</pubDate><guid>https://danhdav.github.io/projects/summer/</guid><description>&lt;p>My university&amp;rsquo;s Association of Computing and Machinery chapter has a &amp;ldquo;&lt;a href="https://www.acmutd.co/projects">Projects&lt;/a>&amp;rdquo; division for students interested in application development every semester. Each project group contains an industry mentor and a project manager for help.&lt;/p>
&lt;p>I was part of a team of four that developed a proof-of-concept study tool app named &lt;a href="https://github.com/acm-projects/Summer">Summer&lt;/a>. This was around the time OpenAI released their first models to the public, so we utilized GPT-3.5 to generate summaries and quizzes from educational YouTube videos.&lt;/p></description></item></channel></rss>