Staff Data Platform Engineer · Data infrastructure, reliability, and production ML

I design and operate data platforms for demanding production systems.

Staff Data Platform Engineer specializing in high-volume ingestion, distributed processing, governance, and hybrid architecture—with production ML engineering across the same lifecycle.

São Paulo, BrazilOpen to selected Senior and Staff opportunities

Discuss a role or platform challenge LinkedIn

Selected production outcomes

Measured improvements across platform reliability and performance.

Each result reflects delivered production work, not a portfolio simulation.

01Martelo

End-to-end Databricks platform

Built the platform across ingestion, transformation, governance, observability, data quality, and production-model delivery.

Model-serving SLO: 99.9%
Data-quality incidents: −40%

02Meta · Processing performance

Hybrid real-time and batch optimization

Removed ingestion and execution bottlenecks across workloads spanning on-premises infrastructure and cloud services.

Real-time processing: +33%
Batch processing: +26%

03Meta · Hybrid estate

Financial-data migration and retained-estate optimization

Delivered two separate initiatives: migrated 40% of high-volume financial-operations data, then optimized workloads intentionally retained on-premises.

Data estate migrated: 40%
Retained on-prem processing: +17%

04Mercado Pago

Critical financial data and regulatory reporting

Improved Snowflake and PySpark delivery across PIX, conciliation, CDB, executive KPIs, regulatory reporting, and the enterprise warehouse.

Data-quality incidents: −22%
Report-creation speed: +25%

Experience

Experience across production data and ML systems.

Platform scope expanded from banking analytics to critical financial data, hybrid estates, production ML, and end-to-end platform ownership.

Oct 2025 — Present

Staff Platform Engineer

MarteloRemote · Florida, United States

Own the architecture and operation of a Databricks platform spanning ingestion, transformation, governance, observability, data quality, and model serving.

Jan 2025 — Oct 2025

Principal Data & ML Engineer

MetaHybrid · Brazil

Led distinct migration and performance programs across a hybrid estate spanning on-premises infrastructure, AWS, and GCP.

Nov 2023 — Dec 2024

Staff Machine Learning Engineer

Plural SolutionsHybrid · Brazil

Engineered production ML workflows from data preparation and feature design through experimentation, deployment, monitoring, and model improvement.

Jul 2021 — Mar 2023

Senior Data Engineer

Mercado Livre · Mercado PagoRemote · São Paulo, Brazil

Delivered critical financial-data products and enterprise warehouse capabilities across Snowflake, PySpark, Kafka, and regulated reporting workflows.

Nov 2019 — Jul 2021

Data Analyst

Itaú UnibancoOn site · São Paulo, Brazil

Automated ETL and analytical workflows while improving customer-data reliability, Snowflake performance, and ingestion efficiency in banking.

Technical capabilities

Data platform engineering, with production ML depth.

Concepts and operating responsibilities come first. Tools follow the architecture, workload, and organizational constraints.

01SourcesApplications · events · databases

02Ingestion & processingBatch · streaming · CDC

03Storage & modelingLakehouse · warehouse · semantic models

04Data productsContracts · APIs · governed datasets

05Analytics & MLDecisions · features · models · serving

GovernanceQualitySecurityObservabilitySLOs & recovery

Primary specialization · 70%

Data Platform Engineering

Adjacent production depth · 30%

ML Engineering & MLOps

Operating principles

How I approach platform ownership.

Architecture decisions are evaluated through production behavior: reliability, change safety, team autonomy, and lifecycle cost.

Design the operating model with the architecture

Ownership, SLOs, recovery, cost, and change control are part of the system—not follow-up work.

Optimize across boundaries

A faster component is not an improvement when it creates instability, opaque handoffs, or downstream debt.

Productize the reliable path

Standards, automation, and self-service should make the correct implementation the easiest one to repeat.

Education

B.Sc. in Computer Engineering — Universidade Presbiteriana Mackenzie, 2021

Electronics Technician — SENAI São Paulo, 2015

Selected certifications

Data Architect Nanodegree — Udacity

Data Product Manager Nanodegree — Udacity

Data Streaming Nanodegree — Udacity

Machine Learning DevOps Engineer Nanodegree — Udacity

Machine Learning Specialization — DeepLearning.AI

Machine Learning Engineering for Production (MLOps) — DeepLearning.AI

Data Engineering with AWS Nanodegree — Udacity

Contact

Open to selected Senior and Staff opportunities.

If the role involves data-platform ownership, hybrid systems, high-volume processing, governance, or production ML, send the mandate and context. I review each request before scheduling a conversation.

Direct details

guilherme_s.ferreira@outlook.com