Founder
Researcher.

Forward Deployed Engineer & Partner · AI agents control plane for corporate benefits

VC-backed startup — invested by Antler, Funses and the angel funds Anjos do Brasil and FEA Angels. Control plane for AI agents across the operations and distribution layer of corporate benefits. I build AI-native and hands-on: I ship and deploy features to production, own the agent control layer — what each agent can run and how it is monitored — and carry the implementations end to end next to the operation itself, turning manual routines — beneficiary movements, eligibility exclusions, carrier layout exports — into orchestrated pipelines. TypeScript/Express REST API over PostgreSQL with TypeORM, a React admin SPA, and Airflow DAGs driving extraction and export.

Co-Founding · Events Ecosystem Platform

B2B2C platform for the live events industry. Unifies ticketing, cashless payments, parking management, stock control, consuming-software and behavioral analytics into a single ecosystem spanning a producer app, ticket validation, and consumer-facing app.

Founder & iOS Engineer · Bill reminders that insist

Live on the App Store. A bill tracker built around the one thing that actually works: reminders that don't give up. Bills heat up visually as the due date approaches, and 'insufferable mode' nags on a decaying cadence through the day until you pay — designed around the iOS 64-pending-notification ceiling. It also writes the awkward 'you owe me' message with your Pix key and opens WhatsApp for you.

Founder & iOS Engineer · Finance with AI

iOS finance app where AI handles the tedious part of staying in control. The dashboard opens on the number that matters — what's left to spend this month. A CoreML classifier categorises transactions on its own, scanning an NF-e QR code or photographing a receipt splits the items automatically, and OFX/CSV import pulls in whole bank statements with dedupe. On top of that: goals with a deadline and a weekly amount, per-category caps that warn before the red, bill reminders, and four simulators — compound interest, the true cost of a cut, debt payoff and installment purchases. All the intelligence runs on the iPhone, so transactions never leave the device.

Personal Project · Small Language Model built from the ground up

End-to-end SLM implementation based on Dr. Raj Dandekar's (MIT PhD, ML) teachings. Covers the full pipeline: tokenization, input–target pair construction, model architecture design, pre-training setup and execution, and inference. Also explores the theory behind it: Backpropagation, activation functions (ReLU & GeLU), and the step-by-step reasoning behind each architectural decision.

Full-Stack Social Network in C++

Social network built from scratch in C++11 with SQLite3, using OOP principles and MVC architecture. Features real-time score ranking, bidirectional follow relationships, and gamification mechanics with robust exception handling.

Space Invaders Game in C

Space Invaders game built from scratch in C, using ALLEGRO library. Features classic gameplay mechanics, enemy movement patterns, and power-up system.

FutureLab · Real Estate Lead Scoring

XGBoost model for real estate profile scoring using Target Encoding, Feature Engineering, and 5-fold Cross Validation. This project was for Morada AI, the model was part of a bigger pipeline of lead scoring and was used to predict the profile of real estate leads based on their messages, behavior and attributes.

From scratch · Neural network implementation

Neural network built from scratch with Pytorch for real estate lead profile scoring.

FutureLab · Research · Knowledge Distillation

Bibliometric analysis of 12,209 scientific publications on LLM and Knowledge Distillation, sourced from Scopus. Built co-occurrence networks in VOSviewer and plotted publication trends in Python — mapping where the research field is heading and what emerging terms are gaining traction.