# About Angel Ortega-Melton

Software engineer building AI-integrated applications, agent tooling, and data pipelines.

## Who I Am

I originally went to UNR to study mechanical engineering after discovering how much I enjoyed my AP physics class in high school. I loved every second of my engineering classes: problem solving, the math (which I used to hate), and the process of understanding how complex systems worked and how to apply them to real life.

Then the pandemic hit. I felt lost and alone, and spiraled out of college.

For a while, software engineering remained something I wanted to pursue but kept putting off. I had always been interested in programming, but I knew people who had been coding since they were kids, and I convinced myself I was already too far behind.

In 2024, I finally decided that didn't matter. I started from the beginning: HTML, then Python, TypeScript, APIs, databases, and the fundamentals of building software. At first, progress felt slow. I knew I could understand technical concepts, but I hadn't found a way of learning that matched how I naturally think.

Then I heard about AI. I started using AI as a learning tool. That changed everything.

For me, it felt like being able to talk to a future version of myself, someone who could explain a concept from five different angles, answer every follow-up question, and immediately help me apply what I had learned.

Not long after, I discovered AI-native development tools like Cursor and began working with AI directly inside my development environment. That was the point where software really clicked for me. Suddenly, I could take an idea and turn it into something real.

Instead of learning concepts in isolation, I could learn by building. Just like in the engineering classes I fell in love with, I could take these abstract things and see them work in realtime. I could discover an API and immediately integrate it. When I encountered an unfamiliar architecture or engineering concept, I could understand it in the context of a system I was actively working on.

That feedback loop became addictive.

Since then, I've spent most of my time building, experimenting, breaking things, rebuilding them, and continually pushing toward more ambitious projects. My interests have increasingly moved toward AI engineering, agentic systems, structured data, developer tooling, and designing systems that make LLMs more reliable and useful.

What started as learning how to write code became an interest in a much bigger question: How can we use AI to make individuals capable of building things that previously required entire teams? That's the problem space I want to keep exploring.

### What I Do

#### AI-Integrated Applications

Full-stack apps with AI at the core, from privacy-focused marketplaces on Sui to investment research tools. Every AI claim carries provenance and quality checks.

#### Agent Tooling

Tool-using agents with structured outputs and prompt pipelines, built on the OpenAI API and LangChain. Policy gates and human review keep autonomous behavior trustworthy.

#### Data Pipelines

Point-in-time datasets from SEC EDGAR and FINRA sources, versioned as JSON artifacts and served through DuckDB over Parquet. Offline pytest suites with pyrefly and ruff gates.

#### Full-Stack Web

TypeScript and React front ends with Hono APIs, PostgreSQL, and Dockerized workers. PWAs and Astro or Vite sites that ship fast and stay maintainable.

### Technologies I Work With

- Python
- TypeScript
- JavaScript
- Rust
- Sui Move
- React
- Vite
- Tailwind CSS
- Hono
- Astro
- OpenAI API
- LangChain
- DuckDB
- Parquet
- pytest
- Docker
- Bun
- Sui
- Walrus
- Seal
- Nautilus

## Professional Journey

### Lead Software Engineer (Freelance)

Apr 2025 - Feb 2026. Dreamlit Apps LLC, Remote.

Built a full-stack dApp environment with TypeScript, React, Sui Move, Walrus storage, and AI integration. Shipped the S.O.N.A.R. audio marketplace end to end as sole creator.

### Warehouse 3rd Shift Supervisor

Mar 2020 - Dec 2023. Amazon, Reno, NV.

Coordinated overnight operations and built inventory-routing workflows that cut errors. Led shift crews through high-volume deadlines with clear handoffs and accountability.

### Key Achievements

- Best Technical Implementation, Walrus Haulout Hackathon 2025, one of three winners from hundreds of entries across 27 countries
- Sole creator of S.O.N.A.R., a privacy-focused audio marketplace on Sui
- Shipped open-source Stockbot with verifiable, provenance-backed data
- Built inventory-routing workflows at Amazon that cut errors

## Get In Touch

I'm always interested in discussing AI-integrated applications, agent tooling, data pipelines, and open-source collaboration.

- [aortegamelton@gmail.com](mailto:aortegamelton@gmail.com)
- [GitHub](https://github.com/aortegamel)
- [LinkedIn](https://www.linkedin.com/in/angel-ortega-melton-31647a186/)

### Let's Collaborate

I'm particularly interested in projects involving:

- AI-integrated applications
- Agent tooling
- Data pipelines
- Open-source AI and developer tools