Looking for a job

José Henrique User

@joshenriqueuser

Message

Data Engineer at Econodata who built Airflow pipelines processing 30M+ images with OCR and data-quality scoring.

Brazil

Message

At Econodata, I architected and deployed an end-to-end Airflow ETL pipeline processing 30M+ images, using OCR for company-name matching and Shannon-entropy scoring for data-quality selection.

I also built web-scraping systems for complex datasets and designed retrieval and graph systems with Elasticsearch and Neo4j to support Web RAG and Graph RAG.

At UFRGS, I researched NLP and built a project tracing AI coding agents at the syscall level with eBPF and strace. I’m currently pursuing an MSc in Computer Science at UFRGS, researching observability and security of AI agents at the kernel level.

Experience

Work history, roles, and key accomplishments

Current

Data Engineer

Current

Econodata

Nov 2025 - Present (11 months)

Architected and deployed an end-to-end Airflow ETL pipeline processing 30M+ images with OCR and Shannon-entropy scoring. Built scalable web-scraping systems and designed a multi-agent retrieval system over Elasticsearch.

GCP Airflow Elasticsearch Neo4j Cloud SQL Lang Chain Langsmith Python Grafana ETL OCR Web Scraping Graph RAG

Full Stack AI Developer

Altervision

Feb 2025 - Nov 2025 (9 months)

Built a serverless data backend on AWS Lambda and DynamoDB serving 4,000+ monthly active clients. Developed automated data-integration pipelines and cut API data consumption by ~40% with complexity-based request routing.

AWS Python AWS CDK PostgreSQL TypeScript REST APIs Serverless DynamoDB Lambda

NLP Research Fellow

UFRGS

Jun 2023 - Feb 2025 (1 year 8 months)

Designed PostgreSQL schemas for training datasets and user logs. Processed large text datasets to train classifiers, reaching 85% F1 in error-pattern identification, and containerized the backend with Docker.

Python PostgreSQL Docker Flask JavaScript NLP Random Forest RNN