Available for new projects

I turn ideas into products that work.

I'm a digital product builder. I take a real problem and ship the complete solution — from zero to production — using AI as my engineering team. Two of these products are running right now, with real users.

See projects → Get in touch
01 — 02Products in production

What I've built

These aren't weekend prototypes. They're systems that run every day, serving real people.

01
Live & in production Automation · Conversational AI

Rafa — WhatsApp virtual assistant

An AI scheduling system for a medical aesthetics clinic

“Rafa” chats with patients on WhatsApp in natural language and runs the calendar on her own: she books, cancels and reschedules appointments, checking real-time availability in Google Calendar. It comes with an admin panel and a financial manager — both installable as an app (PWA). It runs 24/7 serving real patients, on zero-cost infrastructure.

+ Understands natural-language requests (Google Gemini) and books straight into Google Calendar
+ Admin panel + financial manager, installable as an app (PWA)
+ Automatic reminders, daily summary and layered backups with failure alerts
+ Deployed on Docker in the cloud, exposed via Cloudflare Tunnel — $0 infrastructure
Node.jsExpressGoogle GeminiGoogle Calendar APISQLiteDockerCloudflare TunnelOracle CloudPWA
🔒 Admin panel and financial manager are token-protected (patient data) — demo on request
Support panel: Rafa replying to a patient on WhatsApp in natural language about botox, with the patient's personal data redacted.
Fig. 01 — Rafa handling a WhatsApp chat in natural language (patient data redacted).
Architecture diagram of the Rafa bot: patient → WhatsApp Cloud API → Node.js bot, integrated with Google Gemini, Google Calendar and SQLite, with an admin panel and financial manager as a PWA, running on Docker on Oracle Cloud via Cloudflare Tunnel at zero cost.
Fig. 02 — How the pieces connect, from patient to booking.
Clinic financial manager: monthly goal, revenue, expenses, profit, average ticket and charts of revenue vs. expenses and revenue by procedure.
Fig. 03 — Financial manager: goals, revenue and expenses in real time.
02
Live & in production Product · Web app

Planerio — Field sales planner

An app for sales reps to manage their day on the road

A mobile-first app where the rep logs clients, visits, orders and expenses straight from the field. It calculates commissions and monthly profit automatically, flags which clients are “going cold” (no visit in over 30 days) and works even without internet. To save typing, it extracts order data automatically from the quote PDF, using AI.

+ Automatic order extraction from a PDF with AI (Claude)
+ Commission and monthly-profit calculation, plus inactive-client alerts
+ Works offline with haptic feedback — built for use on the road
+ Secure login and cloud data (Supabase), with automated tests
Next.js 16React 19TypeScriptSupabaseTailwindClaude APIVercel
Open the app ↗
Planerio home screen: monthly profit, shortcuts to add a client, log a visit, record a sale and note an expense, visit schedule and a cooling-clients alert. Planerio sales screen: new sale entry with PDF import, commission folders per brand and a list of sales with status and calculated commission.
Fig. 03 — Home dashboard and sales entry (with order import from PDF).
2products live,
with real users
24/7running in
production
$0infra on the
clinic bot
2AIs integrated
(Gemini + Claude)
§About

How I work

I'm a builder: my strength is taking products off the drawing board and making them actually work. Instead of writing every line of code by hand, I use AI as an engineering tool — I define the architecture, guide the technical decisions, test, fix and take it all the way to production.

That makes me fast where it matters: understanding the problem, designing the solution and shipping something people use. The two projects here aren't exercises — they're live systems, with real scheduling, real money and backups I have to make sure never fail silently.

I'm looking for opportunities where shipping value fast matters more than following the traditional path — automation, product and on-demand projects.

What I work with

AI / LLMsAutomationWhatsApp APINode.jsNext.js / ReactSupabase / SQLiteDockerGoogle APIsPWADeploy / Cloud
→Contact

Let's build something together?

If you need someone who ships working products — and fast — I'd be glad to talk.

gbl.andr@gmail.com LinkedIn ↗ GitHub ↗ Download CV ↓