bonolive ~/portfolio
PT
visitor@bonolive: ~/portfolio

visitor@bonolive:~$ whoami

Bonifácio de Oliveira

visitor@bonolive:~$ cat role.txt

Backend & AI Engineer / Curitiba, Brazil · Remote-first

visitor@bonolive:~$ cat headline.md

Backend systems, automation & AI tools — built to ship.

I build practical software: APIs, automation workflows and RAG/LLM systems — from healthcare prior-auth automation to local-AI tooling — that quietly make teams faster.

visitor@bonolive:~$ ls ./actions

View projects Download CV Get in touch

cat about.md

Pragmatic software, shipped from the terminal

I'm a backend / platform engineer with 7+ years building scalable APIs, automation systems and AI-driven tools in Python — services that hold up in production, not just demos.

Most recently, at Thoughtful AI, I built healthcare prior-authorization automations processing ~100 prior auths/day at ~80% acceptance, supporting an estimated USD 600K ROI. My stack runs from FastAPI, PostgreSQL and SQLAlchemy to Docker, Jenkins and multi-cloud (AWS, GCP, Azure), with a recent focus on RAG/LLM tooling, RPA/OCR automation and FHIR healthcare integrations.

ls stack/

Tech stack

Backend

  • Python
  • FastAPI
  • Flask
  • Django
  • Pydantic
  • SQLAlchemy
  • Alembic
  • REST APIs
  • Microservices
  • FHIR

Frontend

  • TypeScript
  • React
  • Svelte
  • JavaScript
  • Astro

Database & Messaging

  • PostgreSQL
  • pgvector
  • Redis
  • RabbitMQ
  • SQL

DevOps & Cloud

  • Docker
  • GitHub Actions
  • Jenkins
  • CI/CD
  • AWS (EC2, RDS, Lambda)
  • Google Cloud
  • Azure
  • OpenStack
  • k3s

AI & Automation

  • RAG
  • OpenAI
  • Ollama
  • llama.cpp
  • PyTorch
  • TensorFlow
  • Keras
  • OpenCV
  • Computer Vision
  • UIAutomator
  • RPA / OCR

find projects -featured

Featured projects

RAG Knowledge Assistant

A self-hostable assistant that answers questions over internal docs with citations — OpenAI in the cloud, Ollama on-prem.

Problem
Teams kept re-asking the same questions because answers were buried across wikis, PDFs and chat history.
Solution
A retrieval-augmented pipeline over PostgreSQL/pgvector with a FastAPI service, swappable between OpenAI and local Ollama models.

Cut time-to-answer for common internal questions from minutes of searching to a single grounded reply with sources.

  • Python
  • FastAPI
  • PostgreSQL
  • pgvector
  • OpenAI
  • Ollama
  • Redis
Read case study

AI-Driven Test Automation Framework

A framework that generates and runs UI test cases automatically, using computer vision to drive real Android devices.

Problem
Manual regression testing was slow, repetitive and the bottleneck before every release.
Solution
A Python framework that auto-generates test cases and validates UI with computer vision + UIAutomator, orchestrated with Docker and Jenkins.

Automated 250+ test cases, saved 100+ developer hours and cut manual testing time by 40%.

  • Python
  • Computer Vision
  • Docker
  • Jenkins
  • RabbitMQ
  • UIAutomator
Read case study

Multi-tenant Sales Platform

A scalable sales-management backend serving multiple clients from one deployment on AWS.

Problem
A growing sales operation needed one system to serve many clients without their data ever crossing streams.
Solution
A FastAPI + PostgreSQL backend with tenant isolation, clean migrations and managed AWS infrastructure.

Supported multiple tenants on shared infrastructure with safe, reproducible schema migrations.

  • Python
  • FastAPI
  • PostgreSQL
  • SQLAlchemy
  • Alembic
  • AWS
Read case study

cat experience.log

Experience

  1. Forward Deployed Engineer

    May 2025 – Apr 2026

    Thoughtful AI Austin, TX, USA · Remote

    • Built and deployed healthcare prior-authorization automation workflows tailored to client operations, supporting ~100 prior authorizations per day in production.
    • Reached ~80% automation acceptance in the initial rollout by analyzing outcomes, identifying failure patterns, and turning operational feedback into improved automation behavior.
    • Supported an estimated USD 600K ROI by automating manual prior-auth processes across payer validation, documentation, and submission workflows.
    • Integrated client systems through APIs, RPA, OCR and workflow automation, enabling agents to retrieve, process and validate healthcare documentation.
    • Ran technical discovery with managers and client stakeholders to map requirements, clarify edge cases, and align automation behavior with real-world operations.

  2. Software Developer

    Aug 2024 – Apr 2025

    Fitec — Foundation for Technological Innovations Manaus, AM, Brazil

    • Built and maintained core backend for a B2B CRM and sales-management platform with FastAPI, PostgreSQL, SQLAlchemy and Alembic.
    • Developed RESTful APIs and database-backed services to centralize commercial data and support internal sales operations.
    • Managed AWS services (EC2, RDS, Lambda) backing cloud-hosted production infrastructure.
    • Designed PostgreSQL schemas and Alembic migration workflows for safer database evolution and lower delivery risk.

  3. Software Engineer

    Feb 2021 – Jul 2024

    INDT — Institute of Technological Development Manaus, AM, Brazil

    • Built an AI-powered Android test-generation framework (Python, LLMs, prompt engineering, UIAutomator) that turned natural-language descriptions into automated tests, lifting acceptance from ~40% to 95%.
    • Automated Android testing with Python, UIAutomator, OpenCV, Docker and Jenkins, cutting manual validation effort and improving QA reliability.
    • Improved the AI generation pipeline by preprocessing test descriptions and filtering noisy code artifacts for more accurate scripts.
    • Ran local and cloud AI experimentation on GCP with dataset versioning and llama.cpp for local-inference optimization.
    • Built a JetBrains IDE plugin in Java integrating internal APIs, and collaborated with Motorola QA to validate and improve generated tests.

  4. Undergraduate Researcher · Elgin R&D Project

    Aug 2021 – Jul 2023

    LUDUS — Laboratory of Technology, Innovation & Creative Economy Manaus, AM, Brazil

    • Built cross-platform thermal-printer tooling with .NET and Python for reliable printer-API communication on Windows and Linux.
    • Designed an automated validation framework for thermal printers, improving test consistency and reducing manual verification effort.
    • Deployed a Flask remote-testing service on Azure to run, monitor and scale printer-validation workflows in the cloud.
    • Prototyped an AI-powered retail recommendation engine for personalized product suggestions.

  5. Undergraduate Researcher · Samsung R&D Project

    Feb 2019 – Sep 2021

    LUDUS — Laboratory of Technology, Innovation & Creative Economy Manaus, AM, Brazil

    • Built a real-time facial-recognition system with Python, Flask, PostgreSQL, TensorFlow, PyTorch and Keras.
    • Designed image-processing workflows to normalize inputs and improve recognition reliability across real-world capture.
    • Developed REST APIs and a React web platform to manage identities, monitor recognition events and expose inference results.
    • Implemented real-time video streaming for continuous facial analysis across multiple users and camera feeds.

  6. Undergraduate Researcher · Scientific Initiation

    Aug 2018 – Feb 2019

    LSI — Laboratory of Intelligent Systems Manaus, AM, Brazil

    • Built an AI recommendation system for e-commerce using Python and Scikit-learn to personalize product suggestions.
    • Improved model performance through metric extraction, experimental evaluation and hyperparameter tuning.

cat education.txt

Education

B.Sc. Computer Engineering

2016 – 2024

UEA — Amazonas State University · Manaus, AM, Brazil

Embedded facial-recognition system using similarity metrics

Nov 2019

Published at the VII Regional Engineering Meeting

Python, similarity metrics and Raspberry Pi.

open contact/

Let's talk

Got a problem worth solving? I'm a message away.

Download CV