Open to opportunities

Gustavo Souza

@gustavosouza

Senior Data Platform Engineer with 8+ years building large-scale data systems. I specialize in streaming pipelines and ML infrastructure.

Brazil

Message

What I'm looking for

I’m looking for a role where I can own and scale data platforms for streaming + ML use cases—turning performance and cost insights into reliable, cloud-native infrastructure that reduces operational load for engineering teams.

I’m a Senior Data Platform Engineer with 8+ years building large-scale data systems from first principles to production. At iFood, I worked on the company-wide ML Feature Store—serving 17+ squads and powering 200+ concurrent Spark jobs processing petabytes of data daily, with online serving latency under 20ms p99. I also led platform improvements that reduced costs, including a 30%+ cut in Redis infrastructure expenses through better resource allocation and pipeline efficiency.

At McGraw Hill, I own the Databricks platform end-to-end and deliver cloud-native, streaming-focused infrastructure that scales reliably while reducing operational burden for engineering teams. I built Cluster Autopilot to optimize job configurations via the Databricks Jobs API (saving over 30%), and created a Spark Performance Agent that ranks worst jobs and uses an LLM to generate code-level recommendations delivered to owners via Slack. Earlier, I re-architected Airflow on AWS ECS Fargate to eliminate weekly crashes and cut infrastructure costs by 60%+, and built CDC and streaming pipelines with Kafka/Kinesis, Spark Structured Streaming, and modern lakehouse tooling.

Experience

Work history, roles, and key accomplishments

Current

Senior Data Platform Engineer

Current

McGraw Hill

Jan 2025 - Present (1 year 6 months)

Own the Databricks platform across the organization, including workspace administration, permissions, access controls, and governance. Built automated cluster optimization and a Spark performance agent to reduce Databricks costs and provide code-level recommendations for poorly performing jobs.

Databricks Apache Spark Cost Optimization LLM Platform Governance Permissions Access Control

Senior Machine Learning Engineer

iFood

Dec 2020 - Jan 2025 (4 years 1 month)

Worked on the company-wide ML Feature Store, serving 17+ squads and powering 200+ concurrent Spark jobs processing petabytes of data daily. Designed streaming and batch feature computation tiers and built an online serving layer with sub-20ms p99 latency.

Scala Apache Spark Databricks Kafka Redis Go Apache Flink Spark Structured Streaming RocksDB Exactly Once Recovery Delta Lake Kubernetes IaC

Data Engineer

MadeiraMadeira

Nov 2019 - Dec 2020 (1 year 1 month)

Re-architected Airflow to run as containerized tasks on AWS ECS Fargate, cutting infrastructure costs and eliminating weekly crashes. Built CDC streaming pipelines to replicate SQL and MongoDB changes into a data lake in near real-time.

Airflow AWS ECS Fargate AWS YAML CDC Debezium Apache Spark Apache Hudi Data Lake MongoDB AWS Lambda KSQL Spark Structured Streaming

Junior Data Engineer

MadeiraMadeira

Dec 2018 - Nov 2019 (11 months)

Maintained and created Scala and Python ETL pipelines orchestrated by Apache Airflow. Worked on data modeling and exploration with Apache Spark and helped build the company’s first Data Lake while monitoring systems with Grafana.

Scala Python Airflow Apache Spark Data Modeling ETL Data Lake Grafana

Research Intern

Positivo University

Mar 2018 - Dec 2018 (9 months)

Conducted research on automated recognition of facial pain expressions in Wistar rats using computer vision and deep learning. Built supporting desktop tooling in C++ with Qt.

Computer Vision Deep Learning C++Desktop Tools