Internal Reporting System — QR-Based Equipment Reports
The client generated equipment reports and technical spec sheets as PDFs sent by email. These were typically viewed once, then buried in inboxes — when a technician or client needed a report again, they had to track down the original email or call the office to resend it, generating 4-8 repeat requests per week.
I built a QR-based verification system. Each piece of equipment gets a QR code linking directly to its current reports and spec sheets — technicians and clients scan it and pull up the information instantly, on-site, with no email search and no phone calls.
Since launch: repeat requests dropped from 4-8/week to zero, the client now charges a premium for the service and cites the panel as a differentiator against competitors, and reports an estimated 30% increase in quote inquiries.
Built solo — architecture, frontend, backend, and deployment.
0
13
Community Platform for a Fan Group — Live in Production
Fans relied on scattered social media posts to stay updated, and new chants or match-day announcements often ended up being spread informally inside the stadium itself — no central, reliable source of information.
I designed and built a community platform end-to-end, solo: architecture, frontend, backend, and deployment. It's live in production with real, ongoing traffic. The admin panel is currently in active development.
Current numbers: 615 visits/week, 4,950 page views/week, 97/100 experience score (Vercel Speed Insights).
0
21
AI-Powered Expense & Inventory Tracker via Telegram
Built a Telegram bot that turns natural-language messages into structured financial and inventory records. A user sends something like "spent 5 on bread" or "sold 5 bags of coffee," and the bot uses Claude's API to classify the message and log it automatically to Google Sheets — no forms, no manual entry.
The system serves two purposes at once: personal expense tracking, and inventory/sales management for a small family coffee business, where stock levels update in real time as sales are logged.
Built solo, end-to-end: bot logic, AI classification pipeline, and Google Sheets integration.
0
26
Landing Page — Café Luchita
Designed and built the marketing site for Café Luchita, a family-run coffee brand from Loja, Ecuador. The site guides visitors through the product story — origin, roast level, grind type — with interactive sections (roast comparison, grind-to-brew-time guide) instead of static text blocks, plus a full ordering flow connected to WhatsApp for direct orders.