2
2
Costa Rica
4+ years of experience
I’m a frontend-focused developer and UI/UX designer based in Costa Rica, with experience building digital products, improving workflows, and connecting design with implementation. My work combines React, Angular, JavaScript, HTML/CSS, product design, QA, automation, and cloud infrastructure experience across AWS, Google Cloud, Docker, and Kubernetes. I enjoy turning complex systems into clear, usable interfaces and practical technical solutions. I’m especially interested in products that bridge design, engineering, automation, and scalable workflows.

AMI — Autonomous Marketplace Intelligence — is a multi-agent AI workspace that helps marketplace and e-commerce teams turn fragmented web signals into ranked, explainable business recommendations. Marketplace teams constantly need to decide what to sell, when to restock, how to price, which suppliers to trust, and which product opportunities are worth pursuing. Today, those signals are scattered across competitor marketplaces, supplier catalogs, product reviews, trend data, inventory movement, pricing changes, and availability signals. AMI uses Bright Data as its live web intelligence layer to collect marketplace, supplier, competitor, and trend signals from public web sources. Those signals are classified, normalized, and routed through specialized assistants. The system includes five coordinated agents: Trend Assistant: detects demand patterns, social momentum, and emerging product interest. Competitor Assistant: tracks pricing pressure, discounts, availability, and market saturation. Supplier Assistant: evaluates sourcing feasibility, supplier cost, delivery risk, and product match confidence. Inventory Assistant: analyzes stock posture, margin context, overstock risk, and operational opportunity. AMI Orchestrator: combines all assistant signals into a final recommendation with confidence, risk, evidence, and a suggested next step. Unlike a static dashboard, AMI is built around decision clarity. It does not simply display raw scraped data. It helps users understand what action to prioritize, why it matters, which agents contributed, what evidence supports the recommendation, and what risks or fallback limitations exist. The MVP supports product discovery, stock optimization, and revenue stock opportunity workflows. By combining Bright Data-powered web intelligence with a structured multi-agent decision layer, AMI helps teams detect opportunities earlier, validate sourcing risks, protect margins, and act faster in competitive marketplace environments.
31 May 2026

CloudShift Radar is an AI-powered cloud migration assessment tool for CTOs, technical leads, DevOps engineers, platform engineers, and development teams preparing to move legacy applications to cloud-native infrastructure. Traditional migration assessments are slow, manual, and often miss hidden blockers until late in the process. CloudShift Radar solves this by scanning repository code through secure static analysis, detecting cloud provider dependencies, hardcoded infrastructure, environment gaps, storage patterns, queues, databases, and other migration-sensitive signals. IBM Bob is used as the project's core AI reasoning engine. After local analysis identifies technical findings, Bob evaluates migration risk, explains likely impact, predicts which features may survive the migration, and generates a practical action plan. The product provides a migration readiness score, unified findings view, human review queue, Bob reasoning trace, and exportable reports. The goal is to help teams know what will break before they migrate, reduce discovery time, and make cloud migration planning faster, safer, and clearer for both technical and business stakeholders.
17 May 2026