Rafael Cedillo@rafael_cedillo264
1
Events attended
1
Submissions made
4+ years of experience
About me
I don’t consume information out of curiosity—I use it as raw material to build systems. My way of thinking is closer to a systems architect than a traditional user. I’m not interested in just learning, but in connecting pieces: artificial intelligence, business, human behavior, and process design. I approach everything as part of a larger system that can be optimized, redesigned, and scaled. I don’t accept tools as they are by default. My focus is on reconfiguring them to operate with greater intent and intelligence. That’s why I work on structuring AI not as a reactive assistant, but as a strategic layer capable of questioning, analyzing, and improving decisions before execution. My profile combines three dimensions that rarely come together: analytical thinking, a deep search for meaning, and a strong drive toward execution. This allows me not only to understand problems, but to turn them into solutions with clear direction and real-world application. I’m aware that the real challenge is not learning more, but transforming knowledge into tangible assets. That’s why my current focus is on building systems and frameworks that can be reused, scaled, and applied across different contexts. In this hackathon, my goal is not just to participate, but to design a solution with strategic coherence, real utility, and the potential to grow beyond the event itself.
🤓 Latest Submissions
.png&w=828&q=75)
ATLAS - Forensic Auditor
ATLAS is an agentic AI platform built for financial forensics and regulatory compliance. It orchestrates four specialized agents: a Vision agent that extracts structured data from tax documents via OCR, a Reasoning agent powered by a custom fine-tuned Qwen3-14B (atlas-r3) that detects arithmetic errors, invalid RFCs, and fiscal anomalies, a Validator agent that cross-checks findings against Mexican and US tax law (CFF, LISR, LIVA, FinCEN, CTA), and an Explainer agent that generates human-readable audit reports. The centerpiece is the Regulatory Sandbox — a predictive simulation engine that analyzes a proposed business operation before it executes, surfacing cross-border tax exposure, regulatory timelines, and alternative restructuring scenarios. The model was fine-tuned in two rounds on AMD MI300X, achieving a final eval loss of 0.1016 on a 13,588-example dataset covering Mexican and US fiscal compliance, cross-border transfer pricing, and adversarial red-team scenarios.
10 May 2026
