Skip to content
View gabrafur's full-sized avatar
🏠
Working from home
🏠
Working from home

Block or report gabrafur

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.

Content in all repositories owned by your account will be closed.
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
gabrafur/README.md

Gabriel B. Furlan — Senior Data and Platform Engineer

Connect on LinkedIn Send an email Explore repositories

Vitória, Brazil · Italian / EU citizen
Eligible to work in the European Union · Open to remote and international opportunities

I design, scale, and operate dependable data platforms — from architecture and infrastructure to delivery, observability, and production recovery.

Impact at a glance

22+ countries · 50M+ records/day · 100K+ users impacted
99.9% reliability · ~35% faster runtime · ~20% lower infrastructure cost

Senior Data and Platform Engineer with 10+ years building production data and software systems and 4+ years specializing in Azure Databricks, distributed data platforms, Infrastructure as Code, CI/CD, observability, and reliability engineering.

I am a technical owner of a global personalization platform, with end-to-end responsibility for architecture, deployment strategy, developer experience, model-output delivery, reliability, and production operations.

How I operate

  • Platform ownership: Own architecture, deployment strategy, developer experience, reliability, and production operations across global data and ML workloads.
  • Technical leadership: Mentor engineers, review code and RFCs/design documents, lead architecture reviews, and turn recurring decisions into engineering standards.
  • Reliability engineering: Design observability, idempotency, reconciliation, safe reprocessing, incident response, RCA, and recovery paths for distributed systems.
  • Cross-functional delivery: Partner with business, data science, engineering, and global platform stakeholders across distributed international teams.

Core technologies

Data and ML platforms

Databricks Apache Spark PySpark Delta Lake Snowflake MLOps

Cloud, infrastructure, and software engineering

Microsoft Azure Terraform Python JavaScript Node.js SQL Docker Docker Compose GitHub Actions

Integration, quality, and operations

RabbitMQ Azure Service Bus MQTT New Relic pytest REST APIs

Automation, edge, and local AI

Home Assistant Node-RED MQTT Zigbee Matter Ollama Model Context Protocol

Featured work

Verified end-to-end data engineering platform built with synthetic data and explicit clean-room boundaries, with both credential-free local execution and a live Databricks Free Edition path.

  • Five-task Databricks serverless PySpark Job using Unity Catalog managed Delta tables
  • Failing data-quality gates, deterministic ranking, SQL MERGE, persisted delivery/outbox simulation, and reconciliation
  • Proven idempotent rerun, bounded replay with zero delivery side effects, and 17/17 commit-bound live acceptance checks

Python · PySpark · Spark SQL · Databricks · Unity Catalog · Delta Lake

View repository E2E evidence CI


Portable, fail-closed runtime for bounded local inference through Ollama-compatible models, with explicit contracts for MCP clients and consuming projects.

  • Dependency-free Python MCP stdio server and CLI with deterministic routing and source-anchored log extraction
  • Checksummed, content-addressed releases with atomic activation and recoverable rollback
  • Metadata-only telemetry, disabled-by-default experimental routes, and tightly constrained endpoint recovery

Python · Node.js · MCP · Ollama · NVIDIA RTX · WSL2 · GitHub Actions

View repository Release CI


Continuously operated, self-hosted platform treated as an event-driven distributed system rather than a collection of device automations.

  • Event-driven architecture, integration engineering, and state recovery
  • Versioned infrastructure, observability, guarded upgrades, and disaster recovery
  • Privacy boundaries and a version-pinned integration with the external Local AI RTX runtime

Python · JavaScript · Node-RED · MQTT · Docker Compose · Home Assistant · Zigbee2MQTT · Matter · YAML/Jinja

View repository Validation

All public portfolio projects use synthetic or sanitized data and contain no proprietary employer source code, production data, credentials, or internal business logic.

Career highlights

Anheuser-Busch InBev / BEES · 2025–Present
Technical owner of a global personalization platform serving 22+ countries and processing 50M+ records/day.

Ambev · 2022–2025
Built and operated Azure Databricks platforms, distributed promotion processing for 200K+ monthly campaigns, and optimizations reducing cloud processing costs by ~70%.

Santander Brazil · 2019–2022
Led data engineering for real-time anti-fraud across 1B+ transactions, contributing to an ~80% reduction in fraud losses (~US$200M).

Itaú Unibanco · 2016–2019
Led forecasting, workforce optimization, and real-time routing initiatives generating more than US$15M/year in operational savings.

Education

  • Postgraduate Specialization in Data Engineering — Escola Politécnica da USP
  • Bachelor's in Aerospace Engineering — Universidade Federal do ABC
  • Technical Degree in Electrotechnics — IFES

Current development

  • Preparing for the Databricks Certified Data Engineer Professional examination
  • Deepening expertise in Generative AI, agent engineering, MLOps, data architecture, and platform engineering
  • Portuguese: native · English: B2 professional · Spanish: basic

Opportunities

I am open to Senior or Staff-level roles in Data Engineering, Data Platform Engineering, Data Architecture, or MLOps, especially opportunities involving global platforms, technical leadership, reliability, and complex distributed systems.

Contact Gabriel Furlan

Pinned Loading

  1. customer-engagement-data-platform customer-engagement-data-platform Public

    Production-inspired data engineering platform with PySpark, Delta Lake, Databricks, reliable pipelines, observability and automated testing.

    Python 1

  2. my_smart_home my_smart_home Public template

    Self-hosted, event-driven home automation platform using Home Assistant, Node-RED, MQTT, Zigbee2MQTT, Python and Docker on Raspberry Pi.

    Python 4