Built with AI

From idea to product

A designer who doesn't just design - but ships. Five projects conceived, built, and launched using AI code editors as co-pilot.

5
Projects shipped
2
Platforms
2026
Year
The mindset

Design is not enough

As a product designer, I can identify real problems. But identifying problems without shipping solutions felt incomplete. I wanted to go beyond wireframes and prototypes - I wanted to build.

AI code editors gave me the ability to go from insight to live product in days, not months. Not as a developer - as a builder who uses every tool available. I bring the product thinking, the UX instinct, the business sense. The AI handles the code execution.

These projects are real and functional - from live services solving everyday friction to automated systems processing real-time data. No mockups - real products.

Project 01

ConsigFlow

Conversational hi-fi prototype focused on the fintech/consignado sector

The product

  • - Interactive chat interface for payroll loans (consignado)
  • - Dynamic data collection and real-time offer calculation
  • - Simulates new loans, portabilidade, and refinancing
  • - Fully functional UI prototype built for rapid validation

AI's role

AI Code Editor as Co-pilot

The AI built the entire conversational UI and the underlying logic to simulate chat responses, calculate loan values, and generate comparison cards. I provided the UX flow, copy strategy, financial rules, and visual direction.

consigflow.app
Open prototype
Project 02

Malagueta

Internal continuous discovery hub powered by AI for product teams

The product

  • - Centralizes user feedback and research notes
  • - AI analyzes and clusters insights into actionable themes
  • - Discovery canvas and impact correlation for prioritization
  • - Seamless navigation simulating an enterprise SaaS tool

AI's role

AI Code Editor as Co-pilot

The AI built the application layout, mocked the AI clustering interaction states, and implemented a Single Page Application (SPA) architecture using Vanilla JS. It strictly adhered to my Swiss Design System tokens for styling.

malagueta.app
Open prototype
Project 03

Vaga Simples

WhatsApp-based automated parking management for condominiums

The product

  • - Residents request parking space swaps and rentals via WhatsApp
  • - Auto-matches requests and applies condominium rules
  • - Generates PDF reports for administrators
  • - LGPD compliant, no app download required

AI's role

AI Code Editor as Co-pilot

The AI handled the code - building the WhatsApp bot integration, the matching algorithm, PDF generation, and the landing page. I made the product decisions: what problems to solve, the user flows, the business model, compliance requirements, and the overall experience.

vagasimples.com.br
VagaSimples - WhatsApp parking management
Visit vagasimples.com.br
Project 04

Mala Fácil

Luggage rental service via WhatsApp - affordable travel without owning a suitcase

The product

  • - Daily luggage rentals from R$12.90 via WhatsApp
  • - Free delivery and pickup - zero hassle
  • - 5-step booking process, no app or registration needed
  • - Sanitized premium luggage with 360° wheels

AI's role

AI Code Editor as Co-pilot

The AI built the landing page, the pricing logic, and the WhatsApp integration flow. I defined the value proposition, the rental model, the target audience, delivery logistics, and the overall brand experience.

mala-facil.netlify.app
Mala Fácil - Luggage rental service
Visit mala-facil.netlify.app
Project 05

EV Alert Bot

Telegram bot that monitors sports odds in real time, calculates Expected Value, and sends instant alerts when it finds +EV opportunities

The product

  • - Monitors live sports odds across multiple bookmakers
  • - Calculates Expected Value (+EV) against a probability model
  • - Sends instant Telegram alerts when a meaningful edge is found
  • - Product experiment in betting, data logic, and automation

AI's role

AI Code Editor as Co-pilot

The AI built the odds scraper, the EV calculation engine, and the Telegram alert integration. I defined the betting model, the EV formula, the alert thresholds, and what constitutes a meaningful edge - the product decisions that make the output useful, not just functional.

Telegram · @EVAlertBot · live

⚡ +EV ALERT

Brasileirão · Flamengo x Corinthians

Implied 45.2% · Model 52.1% · Edge +6.9%

⚡ +EV ALERT

Champions League · Real Madrid AH -1

Implied 38.7% · Model 44.3% · Edge +5.6%

View case study
The process

Human + AI

What I do vs what the AI does - a clear division of responsibilities

  1. 01 Spot real problems I identify friction in everyday life - parking conflicts in condominiums, travelers without luggage. AI can't observe the world like this.
  2. 02 Define the product I decide the value proposition, user flows, business model, and scope. The AI receives clear direction, not vague prompts.
  3. 03 Build with AI AI code editors handle the implementation - frontend, backend, integrations. I review, iterate, and steer the output toward quality.
  4. 04 Ship fast From idea to live product in days, not months. The combination of product thinking and AI execution removes the traditional bottleneck.
View my design work Get in touch