Aegis by Voco

Ports close. Cargo stalls. Aegis reroutes it, you just approve the move.

Comment

Aegis by Voco

Honestly, what made me want to build this is that I enjoy finding complexity within real world problems. After researching for complex, real-world enterprise problems that are out of the box I came across this hackathon and said hey i can win this (this is my first hackathon lol) but back to the story….the ones people aren't really aiming to solve, I study YC startups for fun, and YC YT to learn and was like man this sounded like something I could spend the next ten years on. That's just the honest truth.

What Aegis does: it takes the complexity out of shipping goods overseas by boat. Most of the human error in that world happens in the decisions that can make or break a business financially…..(I learned that while a relationship banker dealing with high networth entrepreneurs) so I wanted to make those parts as autonomous as possible while keeping a human in/on the loop. The AI doesn't make any financial decisions. Bet yet it aims to educate the operator on what's blocking, or about to block, their gains(ROI) the risk, the delays, the compliance issues in the cargo/loads they're managing and recommends a intelligent pathway . A human always approves the actual move.

My Codex story: I'm a huge advocate for Anthropic's ecosystem and most comfortable there, so I was nervous about Codex like seriously i'd never touched it, and I could never tell if the hype was real. So I started with cause this hackathon and started first using Codex and GPT-5.6 as my second opinion in my vertical slice Dev kind of like my auditor. Then my Claude Pro membership hit its usage limit, and that pushed me to fully unleash Codex and see if it could really be unbiased and find the gaps and misalignments in my codebase. What came back blew my mind. I hadn't touched an OpenAI product in almost a year, Claude code just has thought me so much so fast and matches my learning style but coming back like this, I'm glad I'm back. I'll be using Anthropic and OpenAI together from now on. That's how I went from being comfortable in one ecosystem to finding a stack that actually helps me chase my mission: a $100B, solo-founder, zero-employee, AI-native service startup.

What's real: the brain is real and runs live the swarm, the scoring, the rejected-route ledger, the approval-gated case lifecycle, the map, 171 tests. The action at the end is simulated on purpose a human keeps the hands. Live cargo tracking and l m❤cloud persistence are the next wire-ups, and I believe there's a real shot at ten paid pilot cohorts in California to make them live.

What I'm proud of: honestly, where I started versus where I'm at in this whole journey and the confidence I carry now as a founder. This is my first hackathon and it won't be my last. My resume literally says "you miss all the shots you don't take," so I'm taking every shot this year on myself and my venture.

Built With

Share this project:

Updates