Data-AI Architect

What’s the opportunity?

We are looking for a Data Architect to provide the technical direction for dLocal’s data and analytics architecture. This role will shape a scalable, governed, and highly consumable data ecosystem across Data & AI and engineering-connecting domain-owned data products, real-time platforms, analytical workloads, machine-learning use cases, and business-facing data consumption.

You will act as a senior technical reference for architecture decisions, translating business and product needs into pragmatic designs that balance scalability, reliability, latency, security, interoperability, developer experience, and total cost of ownership.

What will I be doing?

• Define and evolve enterprise data architectures, evaluate trade-offs, and recommend fit-for-purpose technology patterns across batch, streaming, lakehouse, warehouse, and operational use cases.

• Review and advise other architects on the data aspects of their RFCs, helping ensure consistency with enterprise data principles, governance standards, and architectural direction.

• Lead the adoption of data-mesh principles, including domain-oriented ownership, data as a product, federated computational governance, self-serve platform capabilities, discoverability, quality, and measurable data-product SLAs.

• Establish reference architectures and engineering standards for data products, pipelines, ingestion, storage, processing, orchestration, observability, lineage, security, and access management.

• Provide oversight on operational SLAs, including latency, cost, quality, freshness, reliability, and production performance.

• Design and govern streaming architectures using technologies such as Kafka, Kinesis, Flink, Spark Structured Streaming, and Databricks, supporting use cases from scheduled batch through sub-second real-time processing.

• Define reliable event-processing patterns, including schema and data contracts, schema registries, event-time processing, late-event handling, idempotency, deduplication, replay and reprocessing, dead-letter flows, and freshness SLAs.

• Shape semantic layers and enterprise ontologies that create consistent business meaning across domains, including canonical entities, metrics, dimensions, relationships, business definitions, metadata, lineage, and versioning.

• Establish patterns that allow semantic models to serve analytics, operational applications, machine learning, and AI use cases without creating duplicated or contradictory definitions.

• Guide the evolution of cloud data platforms and lakehouse capabilities, including Databricks, Unity Catalog, Delta/Iceberg tables, object storage, data warehouses, and BI consumption layers across AWS and GCP environments.

• Provide architectural direction for MLOps and feature-platform capabilities, including batch and online features, model-serving integrations, low-latency data paths, model/data lineage, monitoring, and governance.

• Lead or contribute to architecture RFCs, technical decisions, design reviews, migration plans, and implementation roadmaps; make complex trade-offs clear to both technical and non-technical stakeholders.

• Partner with domain teams to clarify ownership, data-product responsibilities, operational handover, quality accountability, access approval, and cross-domain consumption models.

• Define practical controls for data quality, observability, privacy, security, resilience, cost management, and production readiness.

• Take ownership of critical architectural issues, facilitate resolution across teams, and ensure decisions are followed through to implementation and operation.

• Act as a trusted advisor and technical mentor to data engineers, platform teams, data scientists, MLOps engineers, BI teams, and engineering leaders.

• Communicate a cohesive architectural vision while remaining pragmatic, adaptable, and close enough to implementation to validate that designs work in production.

What skills do I need?

• 8-10+ years of experience designing and operating scalable data architectures, preferably in complex enterprise or high-growth environments.

• Strong experience designing and implementing data-mesh architectures and operating models, including domain ownership, data products, federated governance, self-serve platforms, contracts, quality, and discoverability.

• Deep experience with streaming and event-driven architectures, including Kafka or Kinesis and one or more processing engines such as Flink or Spark Structured Streaming.

• Demonstrated ability to design for real-time and near-real-time workloads, including latency measurement, event-time semantics, late data, state, deduplication, idempotency, replay, and failure recovery.

• Strong knowledge of semantic layers, business ontologies, canonical data models, knowledge graphs or metadata models, metric definitions, and semantic governance.

• Expertise in data modeling, data lake and lakehouse patterns, warehouse design, data pipelines, data products, metadata, lineage, and data management technologies.

• Experience with cloud data platforms and services, particularly AWS and/or GCP; experience with Databricks, Unity Catalog, Delta Lake, Iceberg, or comparable technologies is valuable.

• Proficiency with relevant data and platform technologies such as Spark, Airflow, dbt, Kafka, Python, SQL, CI/CD, infrastructure-as-code, and observability tooling.

• Experience architecting or supporting MLOps, feature stores, online/offline data serving, or other low-latency machine-learning data systems.

• Ability to establish practical frameworks for data access, stewardship, governance, privacy, security, quality, and operational accountability.

• Strong understanding of reliability, scalability, performance, resilience, cost, and vendor lock-in trade-offs.

• Excellent stakeholder-management, communication, facilitation, and influencing skills, including the ability to balance delivery expectations and technical excellence.

• Comfortable managing risk, ambiguity, and conflict; able to make decisions and explain the reasoning behind them.

• Self-sufficient and proactive, with the judgment to know when to seek input and when to move forward.