Open to remote roles & freelance with US and European teams

Vagner Bessa

Senior Developer & AI Engineer

I design and ship production-grade AI products and web platforms — Python & TypeScript.

Let's talk See my work

GitHub ↗LinkedIn ↗Download CV ↓bessavagner.dev@gmail.com

I build smart, fast solutions end to end — from AI agents and APIs to the interfaces people use.

Remote · overlapping US & EU business hours · based in Brazil (UTC−3)

Stack & toolbox

Languages

Python
TypeScript
JavaScript
C++
Bash

Backend & AI

FastAPI
Django
aiohttp
Fastify
Mastra (AI agents)
LLM/RAG
XGBoost / ML

Frontend

React
Svelte
Tailwind CSS
DaisyUI
Vite

Data & Infra

PostgreSQL
Prisma
Docker
Google Cloud Run
Prometheus/Grafana
Linux

Selected work

production

Architect & lead developer

Agente Clínicas

AI-powered WhatsApp service for medical clinics

A production-grade monolith that automates patient communication for clinics over WhatsApp: scheduling, triage, emergency detection and clinic management, with full LGPD (Brazilian GDPR) compliance.

Multiple specialized AI agents (assistant, triage, scheduler, emergency)
Full WhatsApp Business integration with multi-instance support
LGPD compliance with audit logging and PII redaction
Resilience: retry with backoff, circuit breakers, rate-limit bypass for emergencies
Full observability stack and multi-clinic data isolation
Role-based React admin dashboard (clinics, patients, doctors, appointments)

TypeScript
Fastify
Prisma
Mastra AI agents
Baileys / WhatsApp
React (SPA)
Redis
Prometheus
Grafana
Sentry

Project details →View code

active

Lead developer / ML & AI engineer

Public-Sector AI R&D · IFCE Innovation Hub

Applied ML & LLM research on public-sector financial data

An applied AI/ML research-and-development engagement at IFCE's Innovation Hub, supported by EMBRAPII, building machine-learning and LLM-based tooling for public-sector financial data. Project specifics are covered by a confidentiality agreement and are not disclosed here.

Applied machine learning and LLM techniques to public-sector financial data
Data engineering and analytical tooling on a Python / Django stack
Conducted within IFCE's EMBRAPII-supported Innovation Hub
Methods, components and results withheld under a confidentiality agreement

Python
Django
Machine learning
LLM-based tooling
RESTful APIs
uv

Project details →Private / under NDA

active

Solo full-stack developer

Expense Tracker AI

Conversational bookkeeping from voice notes and receipt photos

A production Django app for household finance with a PydanticAI assistant that turns plain language, voice notes and photographed receipts into categorized ledger entries — installed on Android and used daily.

Conversational bookkeeping — describe a purchase in plain Portuguese and the assistant resolves category, payment method and billing month before saving
Receipt OCR — photograph a receipt; a vision model extracts line items and Python prorates discounts deterministically so the split always reconciles
Voice notes — record a memo; it is transcribed and routed through the same registration flow, with media processed then discarded
Semantic memory — per-user rules stored as pgvector embeddings and recalled across sessions
Hybrid architecture — HTMX for CRUD pages, React islands for the dashboard and chat, shipped as an installable Android PWA/TWA