I build Python APIs and automations with a focus on predictability and scale. My foundation comes from years inside CRM and MarTech operations: connecting legacy platforms, handling data inconsistencies and keeping critical processes running.
Today I apply that support and engineering background to design backend systems that are ready for the real world from the start — thinking about the database, the infrastructure and the whole flow of information.
years in tech
6
public repositories
contributions in the last year
scroll to explore↓
What I work with
Backend
Python APIs with token authentication and hashed passwords.
Python
FastAPI
Django
REST APIs
JWT with refresh
bcrypt with pepper
Architecture
Layered code, with business rules kept out of the entry layer.
Layered architecture
SOLID
Clean Code
Decoupled services
Data
Relational modelling and queries on PostgreSQL, MySQL and Oracle.
PostgreSQL
MySQL
Oracle
SQL
Relational modeling
Infrastructure and DevOps
Containerised services and Linux deploys behind a reverse proxy.
Docker
Linux (WSL and Ubuntu)
Git and GitHub
systemd
nginx
Proxmox
Testing and quality
Automated tests with pytest and verification by execution.
pytest
API testing
Code review
Verification by execution
Other languages
They come in when the task goes beyond the backend: interface, operations script or automation.
Java
JavaScript
Node.js
React
Bash
Automation and AI
Automated flows with LLMs used as an engineering tool.
Process automation
API integration
Agent orchestration
Multi-agent on Slack
LLM as an engineering tool
Hermes Agent
Learning
What I already use while building projects, with study and feasibility work as I learn.
OAuth2
MongoDB
AWS
Microservices
TDD
UiPath
Projects
monolith · web
Projeto Carros
Web application for registering and managing vehicles, built with Python and Django on the Model-View-Template architecture. Access control and authentication protect the administrative and operational routes. Dynamic filters search vehicles by several criteria, and file upload stores the photos. Persistence and business rules rely on the Django ORM and strictly validated forms.
Architecture
Templates (HTML5/CSS)
Django Views (MVT)
ModelForms (validation)
Django ORM
Database (PostgreSQL)
Highlights
Python and Django backend with the MVT pattern, separating presentation from models and rules.
Native authentication flow that restricts access and protects the editing forms.
Static file upload and management tied to the model lifecycle.
Search and filtering backed by optimised Django ORM queries.
Asynchronous catalogue and shopping intelligence API on FastAPI and PostgreSQL, organised in decoupled service layers. It processes raw data from Brazilian electronic fiscal receipts (NFC-e) into a cross-store price history. Pydantic schemas validate inbound contracts and serialise responses. A pytest suite keeps the data pipeline stable.
Architecture
NFC-e payload
FastAPI Routers
Pydantic Schemas
ETL service
PostgreSQL
Highlights
Layered FastAPI architecture with asynchronous I/O operations.
Dedicated ETL pipeline for normalising, deduplicating and storing NFC-e receipts.
Inbound and outbound contracts typed and validated with Pydantic at the edge.
pytest suite covering endpoints and transformation rules.
public MIT-licensed version; main development in a private repository
rest api · AI
Pimp My Commander
Asynchronous FastAPI application for recommending and building Commander decks in Magic: The Gathering. It integrates the Scryfall, EDHREC and ManaBox APIs to map the player collection and find statistical card synergies. Language models analyse archetypes and justify card swaps with context. Everything runs in Docker containers, which keeps the environment reproducible.
Architecture
FastAPI Routers
ManaBox Parser
Gateways (Scryfall / EDHREC)
LLM engine
Docker
Highlights
Resilient external integration with Scryfall, EDHREC and ManaBox exports.
LLM orchestration to analyse context and recommend card substitutions.
Response contracts validated end to end with Pydantic.
Containerised environment with Docker for local and production runs.
Local portal that centralises technical procedures, publishes reports and tracks operational incidents. Fast search, severity filters and indexing support technical teams in real time. Delivery goes through a reverse proxy on Nginx with access control via basic authentication. The interface uses plain HTML, CSS and JavaScript, avoiding heavy dependencies and prioritising immediate load during incidents.
Architecture
Nginx (proxy and auth)
Interface (HTML / CSS / JS)
Python service
Local parser and logs
Highlights
Edge and security layer on Nginx providing basic user authentication.
Plain JavaScript frontend built for speed during contingency.
Local search and reactive filters to find routines and reports quickly.
Python scripts that read and consolidate incident files.
Customer journeys and marketing automation with segmentation, personalization and campaign rules.
Communication channels: email, SMS, push notifications and social media.
Integrations with Customer Data Platforms (CDP) and CRM systems.
Worked with Marketing, CRM and Technology teams to organise customer data and campaign requirements.
Ads Technical Support Specialist
2023 to Oct 2024
Atento Brasil
Specialized technical support for Meta Ads campaigns.
Campaign performance analysis based on data reports.
Tracking setup with the Meta Pixel.
Resolution of complex technical issues in ads and campaign implementation.
Translation of technical information for non-specialist users.
Courses
Python Master (Logic, Django and Integration)PyCodeBR
Complete package combining four tracks: Python logic and algorithms for beginners, full-stack Django with APIs and AWS deploy, applied AI with LLMs and autonomous agents, and system integration with REST APIs and webhooks.
AI MasterPyCodeBR
AI applications with LangChain and RAG, autonomous agents and chatbots wired to channels such as WhatsApp.
AI Master ElitePyCodeBR
Eight months of live mentoring covering advanced AI deployment, observability, agent architecture and the business side of AI solutions.
AI BuildersPyCodeBR
Architecting and shipping complex AI flows with OpenAI, LangChain, RAG, autonomous agents and MCP. Also covers Claude Code, Codex and Gemini CLI to plan and build systems, plus the product view: turning this into SaaS or an enterprise solution.
Autonomous Agents ImmersionPyCodeBR
Infrastructure first: instead of only writing prompts, the structure around the agent — deploy on a Linux VPS, domain, DNS, Cloudflare and SSH access. The goal is continuous operation, with connected tools, long-term memory and context bases.
FastAPI WorkshopPyCodeBR
High-performance APIs with FastAPI, covering Docker, Kubernetes, tests, CI/CD and scalable deploy on AWS.
Languages
Portuguesenative
Englishfluent
Spanishbasic
Contact
Connect with me on LinkedIn and explore my projects on GitHub.