Skip to content
View carinepra's full-sized avatar

Block or report carinepra

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
carinepra/README.md

Hi, I'm Carine 👋

Senior Full-Stack Engineer · AI-Native 📍 Belo Horizonte, Brazil · Open to remote opportunities worldwide


About me

I build web products that work — fast, accessible, and maintainable. My background is in Control & Automation Engineering (UFMG), which means I think in systems: inputs, outputs, feedback loops, edge cases.

I've spent 7+ years mostly in React and Next.js, but I'm comfortable across the full stack (Python, DynamoDB, REST APIs, and currently Go) and I care deeply about the space where engineering meets product and design.

My relationship with AI in development started with GitHub Copilot for autocomplete and evolved — through real mistakes, structured experimentation, and a lot of iteration — into a fundamental shift in how I work. I've gone from "make it work fast" to "deliver with high quality in the same time", with AI handling the mechanical parts and me focusing on architecture, judgment and product decisions.

In November 2025 I documented and presented this evolution to the entire engineering team — what worked, what failed, and what changed. That same month, I was recognised as the highest AI user in RedVentures' global developer ranking — Cursor reached out directly for product feedback.

The practical result: I built an end-to-end dev pipeline at my current company using state machines, automation scripts and AI rules (Cursor + Claude AI) that takes a task from Jira to an opened PR automatically. Developers across multiple teams have adopted it organically since.


Things I've shipped

Most of my production work is covered by an NDA. The items below are generic, company-agnostic versions of internal tooling, deliberately adapted and published with agreement.

  • Dream AI — a trauma-informed Image Rehearsal Therapy prototype for combat veterans and assault survivors, designed, built, and reasoned about end-to-end with AI. Built over 3 days as a paid design exercise; full process documentation, prompt chain, and decision rationale included.
  • HR platform for Icatu (Onze) — greenfield product, Next.js + Python + DynamoDB, shipped on a hard deadline.
  • AI dev pipeline — 12 scoped Cursor Agent skills + persistent state + optional Jira/GitHub integrations, orchestrating the full delivery lifecycle from ticket to opened PR.
  • Aurora Labs Academy — DynamoDB Quest — gamified quest platform teaching DynamoDB access patterns and single-table design, built under deadline with AI agent orchestration. Built to train an engineering team; high engagement, well-rated by devs and leadership.
  • Pension app Design System (Onze) — internal npm package adopted across 3 of Onze's products.
  • SWR migration — took the main app from a multi-second load + peak-event crashes to near-instant load and the company's first failure-free annual peak event.
  • Python + Selenium RPA — automated a high-volume document workflow integrated with legacy enterprise software for a freelance client; maintained and evolved across multiple cycles.
  • 🏆 1st place — Techstars Startup Weekend Women BH (2026) — built an MVP (lead-capturing landing page wired to a live back-end, plus a front-end prototype) for cuidacomigo.com.br in 54 hours; advanced to the international stage.

Stack at a glance

Front-End   →  React · Next.js · TypeScript · Vue.js · HTML/CSS/SASS · SWR · Storybook
Back-End    →  Python · Go (in progress, AI-assisted) · PHP · REST APIs · Webhooks
Data        →  DynamoDB · MySQL
AI & Tools  →  Cursor (Claude AI) · ChatGPT API · state machines · NotebookLM
Monitoring  →  LogRocket · Grafana · AWS CloudWatch · ArgoCD

A few things that made me who I am

  • Moved to Turkey for an exchange program and, when the planned work fell through, taught myself full-stack development in a large undocumented codebase written entirely in Turkish — with no English-speaking teammates. First production delivery: a few weeks later.
  • 5 years in B2B sales strategy before pivoting to engineering. That background makes me genuinely useful in rooms where product and tech decisions get made together.
  • I've led accessibility guilds, done 1:1s with junior devs, and configured LogRocket dashboards to turn user-session data into product decisions. Engineering for me has always been inseparable from the people using the thing.

Let's connect

LinkedIn Email

Popular repositories Loading

  1. carinepra carinepra Public

  2. ai-dev-pipeline ai-dev-pipeline Public

    AI development pipeline for Cursor Agent — orchestrates the full delivery lifecycle from Jira ticket to open PR using 12 scoped skills, persistent state, and optional GitHub/Jira integrations.

    Python

  3. dynamo-lab dynamo-lab Public

    Gamified quest platform for learning Amazon DynamoDB access patterns and single-table design — built with AI agent orchestration

    JavaScript

  4. dream-ai dream-ai Public

    A trauma-informed Image Rehearsal Therapy prototype for combat veterans and assault survivors — designed, built, and reasoned about end-to-end with AI.

    HTML