> AI PRODUCT ENGINEER
Natan Rivera
I build AI agents and ship full-stack products. Mechanical engineering background from Rice University — I've worked on gigafactories and data centers, now I build software.
Open to product engineer roles and freelance projects — let's build something together.
// ABOUT
Who I Am
Building
6 Products Shipped
AI Agents, Full-Stack Apps, Hardware
Education
Rice University
B.S. Mechanical Eng., Minor in Data Science
Domain Experience
Gigafactories & Data Centers
MEP Asst. PM — Advanced Manufacturing
Pole Vault
National Record Holder
El Salvador
I build AI agents and full-stack products. Rice University mechanical engineering grad with a minor in Data Science — I've managed MEP scope on gigafactories and data centers, and now I channel that same rigor into shipping software.
My focus is on LangGraph-based AI agents, React Native mobile apps, and Next.js web products. I built WORKA (an AI fitness coach with 30 paying users) end-to-end as a solo developer, from the multi-step agent pipeline to the App Store submission. I ship fast and I ship things people pay for.
I also hold El Salvador's national pole vault record — the same obsessive drive that got me there is what I bring to every product I build.
TECH STACK
// PROJECTS
Things I've Built
From AI coaching agents to BLE heart rate sensors on Raspberry Pi — I like building across the full stack.
WORKA
LIVEAll-in-one AI fitness coach app — 30 paying users, live on the App Store. Solo-built from ground up: React Native/Expo mobile app, TypeScript backend, and a multi-step LangGraph agent that handles coaching logic across 4 tool nodes with memory persistence. Wearable integrations with Apple Watch, Apple Health, and BLE sensors.
Pixel Carrousel
LIVEAI-powered carousel slide generator built on a multi-node LangGraph agent pipeline — describe your topic and get polished, brand-consistent slides rendered via Puppeteer. The agent orchestrates content generation, brand guideline synthesis, and layout rendering across separate tool nodes. Features custom font & color palette selection and an image asset store for plug-and-play brand kits.
Noah AI — Second Brain
ARCHIVEDAI-powered note organizer inspired by Cursor but for notes. Built with a LangGraph agentic pipeline, a RAG retrieval layer using Milvus vector embeddings, and a Neo4j knowledge graph that automatically links and clusters related notes for contextual retrieval.
Fit Instinct
LIVECo-founded and co-own this fitness studio brand with my brother. Built the business from the ground up — operations, marketing, sales pipeline, and client acquisition. Currently implementing our in-house BLE Heart Rate Platform for vertical integration and planning expansion.
CuentasClara
LIVEFree personal finance tool for Salvadorans — a budgeting platform for managing income and expenses. Built to solve the lack of accessible money management tools in Central America.
BLE Heart Rate Platform
BUILDINGRaspberry Pi bridge that receives BLE/ANT+ signals from any heart rate sensor and feeds real-time workout data into our own dashboard. Runs on Chromium kiosk mode with Web Audio cues. Full stack ownership from receptor to screen, no third-party platform dependencies.
// PREVIOUS ENGINEERING
Domain Experience
AESC 30GWh Gigafactory — South Carolina — EV battery gigafactory for AESC. Led mechanical scope across a massive advanced manufacturing facility.
Rivian Manufacturing Facility — Normal, Illinois — MEP buildout for Rivian's manufacturing and storage expansion.
Data Center — Confidential — Confidential client — MEP scope for hyperscale data center infrastructure.
// RESEARCH & ACADEMIC
▶Rice University
// RESEARCH & ACADEMIC
▶Research, capstone projects, and published work from my time at Rice.
Nanoscale Heat Transfer Lab — Rice University
GCURs 2021 WinnerCarbon Nanotube Fibers for High-Voltage Conductors
Researched the feasibility of CNT fibers as high-voltage conductors with a technoeconomic analysis. Presented at GCURs 2021 and won. Published in the International Journal of Heat and Mass Transfer. Also co-authored a separate published paper from the lab.
Capstone Project — Houston Fire Department
HFD Motor Vehicle Incident Detection
Built a custom feature engineering pipeline and unsupervised outlier detection model to identify patterns in vehicle incidents. Used Isolation Forest, Local Outlier Factor, and Minimum Covariance methods. Achieved 98%+ accuracy with <2% false positive rate.
Capstone Project — Schlumberger
Deep-Sea Tree Cap Removal Tool
Designed a tree cap removal tool for deep-sea wellhead maintenance. Full mechanical design process — tolerance stack-up, CAD modeling, and FEA stress analysis.
// CONTACT
Let's Build Something
Need AI solutions or custom hardware for your business, have a project in mind, or just want to connect — reach out.