
3
3
Colombia
1 year of experience
I'm a founder and developer based in Colombia, driven by a simple goal: building technology that actually solves hard problems. Whether I'm mapping out complex traffic flows or diving deep into medical AI, I believe the best code is the kind that has a real-world impact. Currently, I'm pouring my heart into OncoTriage, a project designed to help doctors catch lung cancer more accurately. There’s something powerful about turning lines of code into a 'glass cockpit' that can help save lives. I'm a firm believer in building in public, staying grounded in functional results, and pushing the limits of what a small, dedicated team can achieve from a home workstation. When I’m not in the zone with a pair of headphones on, I’m likely looking for the next 'impossible' challenge to tackle.

Reaper Eagle Scout is a public web opportunity intelligence platform built for startups, founders, entrepreneurs, and small vendors who cannot afford to miss tenders, RFPs, grants, and procurement opportunities hidden across the open web. The platform uses Bright Data’s web access ecosystem, especially SERP-style public web discovery, to find live opportunity signals from sources such as public procurement portals, tender databases, government notices, and market-facing web pages. A Python FastAPI backend processes the discovered results through an enrichment and scoring pipeline, while a React frontend presents the results in a command-center dashboard. Because public procurement data is often messy and inconsistent, Reaper Eagle Scout includes an inference layer that enriches incomplete sources using titles, snippets, URLs, known procurement domains, extracted dates, and source reliability. This allows the MVP to infer entities, regions, opportunity type, deadline urgency, confidence, and strategic relevance even when the original page does not provide perfectly structured data. The project is not a generic scraper or chatbot. It is a deployed web-data intelligence workflow: Bright Data-powered discovery, backend enrichment, opportunity classification, urgency estimation, confidence scoring, and a live dashboard hosted with Netlify and Railway. In production, Reaper Eagle Scout can be expanded with deeper portal-specific parsers, document extraction, saved searches, alerts, startup fit profiles, and richer opportunity briefs. The mission is simple: give small teams the kind of opportunity intelligence large enterprises already have, so they can move before the market does.
31 May 2026

TrafficOS is an AI-powered urban traffic intelligence platform designed to transform raw mobility data into operational decisions for cities, logistics teams, and delivery fleets. Instead of simply displaying traffic on a map, TrafficOS analyzes live road conditions, estimates congestion levels, forecasts future traffic density, and explains route decisions through an AI intelligence layer. The system combines Mapbox visualization, TomTom live traffic data, SARIMAX-style forecasting, and a Gemini-powered assistant called ARIA. TrafficOS monitors road segments, estimates speed, density, flow, congestion tier, and route risk, then converts those technical signals into clear operational reports. ARIA allows non-technical users to understand what is happening in the network, why a route is recommended, and what conditions may affect delivery time, fuel efficiency, or congestion exposure. A key part of TrafficOS is its use of exogenous variables in forecasting. The model is designed to account not only for traffic history, but also for weather, precipitation, time of day, day of week, holidays, festivities, incidents, road events, and logistics pressure. This makes the system more realistic for real-world urban mobility, where congestion is rarely caused by one factor alone. The demo supports Fusagasugá, Colombia and San Francisco, USA to show how the same architecture can adapt to both emerging-city mobility challenges and dense metropolitan logistics scenarios. TrafficOS is especially relevant for delivery companies, fleet operators, and urban planners who need to optimize fuel, time, reliability, and dispatch decisions. The platform includes a live traffic dashboard, dynamic routing interface, segment analytics, Gemini/ARIA city pulse intelligence, and PDF report generation. The goal is to move beyond consumer navigation and build a command-center style system for operational traffic intelligence.
19 May 2026

OncoTriage is a clinical decision support system designed to detect and triage lung nodules in chest X-rays with high reliability. Developed solo for the 2026 AMD Hackathon, it addresses the lack of transparency in automated diagnostics by implementing Bayesian Deep Learning. The system utilizes a Bayesian EfficientNet-B4 backbone. By employing MC Dropout, the model generates a predictive distribution rather than a single point estimate, allowing for the calculation of epistemic uncertainty. This effectively quantifies the model's confidence for every detection. In clinical settings, this allows the system to flag low-confidence predictions for priority human review, reducing the risk of false negatives inherent in standard "black-box" AI. In addition to this, in order to handle the intensive computational requirements of Bayesian inference, OncoTriage is optimized for AMD Instinct MI300X instances. Leveraging AMD’s high-bandwidth memory (HBM3) and the ROCm stack, the system achieves the rapid inference times necessary for real-time clinical triage. The environment is fully containerized via Docker, ensuring seamless scalability across high-performance compute clusters. The Mission: OncoTriage represents a shift toward accountable, transparent AI. By bridging the gap between raw computational power and clinical safety, it provides radiologists with a reliable partner in oncological screening—transforming raw data into uncertainty-aware medical intelligence.
10 May 2026