Senior Data Engineer - São Paulo (Contract)

Martelo turns public records into commercial intelligence for the U.S. construction industry. We ingest building permits, deeds, tax assessor records and property data from sources across the country, resolve the conflicts between them, and deliver clean, connected data products that manufacturers, distributors and contractors use to find their next customer.

We started by transforming messy, fragmented property and permit data into clear, actionable insight. That foundation is now live and in customers' hands, and here the data is the product: the quality of our entity resolution, our reconciliation logic and our data models is what our customers pay for.

We are still a small, ambitious team that thrives on curiosity, resilience and purpose. We are builders, crafting technology from the ground up, exploring, experimenting and learning fast. The journey has moved past its first chapter, but the vision is unchanged: to build the leading data-intelligence platform for construction.

Your Title: Senior Data Engineer.

Your Location: Hybrid/Brazil HQ address - Nações Unidas 12901 - São Paulo - SP (the team meets at the office once in a while.).

About the Opportunity:

Senior individual contributor with end-to-end ownership. You gather requirements with go-to-market and product, design the solution, build it, prove it's correct, and keep it healthy in production. You lead the execution of your own work, and we expect you to propose work to the roadmap when you spot it before we do.

Our platform runs on Databricks and you'll be working on it daily from week one. We are not set up to ramp someone on the fundamentals.

What you will do:

• Entity resolution and master data management: source-of-truth rules, survivorship logic, stable identifiers and manual overrides across permits, deeds, assessor records and business/contact data. This is the heart of the role.
• Reconciliation of third-party sources that disagree, at national scale, plus the observability that tells us when the logic drifts.
• Model stewardship: land new sources inside our production data models, and evolve those models when they no longer fit instead of bolting on workarounds.
• Data quality end to end: testing, monitoring, alerting, lineage, and root cause analysis when incidents happen.
• Analysis that shapes the roadmap: which sources to acquire next, which datasets to prioritize, and what quality gaps actually cost us commercially.
• Productionizing data science models: the data models, contracts and drift monitoring around ML.
• Technical reference for the team: code review focused on correctness, cost and maintainability trade-offs.
• AI-assisted engineering. We use AI agents in day-to-day development, and you'll apply them where they add leverage, with guardrails and traceability.

What success looks like in your first year: New sources land in production models without destabilizing them. Entity resolution quality is measured, monitored and visibly improving. At least one recurring data quality problem gets solved at the architecture level instead of patched again.

What you bring:

• Strong SQL and Python, applied across the full data lifecycle from ingestion to serving.
• Solid, hands-on Databricks experience: notebooks, jobs, workflows, Delta Lake and PySpark in production. It's our core platform and we won't be teaching it.
• Data modeling depth: dimensional modeling, slowly changing dimensions, grain design, medallion or equivalent layered architecture.
• Entity resolution and data reconciliation. This is the core of the role.
• Data quality and observability: testing, monitoring, alerting, lineage.
• Transformation and orchestration frameworks (dbt, Airflow, Databricks Workflows or equivalent), with version control and CI/CD applied to data models.
• Experience with data products feeding a live application, not only dashboards, and the ability to translate an ambiguous business question into an actionable number.

It will be a differential if you have:

• Data product architecture. We run multiple data products (permits, deeds, tax assessor, businesses and contacts), each built from several sources. Experience designing decoupled, product-oriented data architecture will set you apart: how domains map to models, how pipelines are organized and run per product, and how data is laid out and stored in a lakehouse.
• AI architectures applied to data: pipelines for RAG, MCP for standardized data-source access by agents, and quality and traceability in non-deterministic agentic systems.
• Advanced Databricks ecosystem: Unity Catalog, Delta Live Tables.
• Data governance: access control, cataloging, lineage.
• MLOps beyond the basics: model serving, drift monitoring, retraining workflows.

Our stack: Python, SQL, PySpark, Databricks, Delta Lake, dbt, AWS, Terraform, Postgres, GitHub Actions, and AI agents in our day-to-day development.

Martelo Benefits:

• Competitive monthly compensation with real impact from day one.
• Paid vacation, sick leave, and health support allowance, because taking care of yourself matters as much as building.
• Local holidays off, so you can recharge with family and friends.
• The chance to work with cutting-edge technologies and tackle data challenges no one has solved before.
• A collaborative, inclusive, and diverse culture where every idea can shape the product and company.
• Fast track professional growth. Learn, experiment, and build skills that will accelerate your career.

If you meet most of what we need and none of the differentials, apply anyway. We'd rather talk to you than miss you.