We're looking for the right person to build that foundation from the ground up. This is a hands-on leadership role with real ownership: you'll shape how data is collected, structured, and delivered — and grow a team around you as the function matures.
WHAT YOU'LL BE WORKING ON
Design the target-state data architecture: Postgres CDC → Kafka → Snowflake → client-facing data products.
Own tooling decisions across ingestion, orchestration, transformation, and quality layers.
Implement CDC-based ingestion from PostgreSQL services (RDS, Aurora, EC2, K8s operator) using Debezium or equivalent.
Build streaming and near-real-time pipelines with defined SLAs.
Build a data quality control layer: checksums, reconciliation, schema validation, anomaly detection, and alerting.
Define quality checkpoints across the full pipeline — from source capture through Snowflake to client delivery.
Define and enforce data contracts with service-owning teams for core entities: transaction, merchant, settlement, and processing status.
Build the external data delivery layer: financial settlement, transaction status, processor reconciliation, and client analytics.
Design tenant separation and implement replay/reload mechanisms for failure recovery.
Start hands-on, then gradually hire and grow a small data engineering team as the function matures. Build a pragmatic roadmap with concrete deliverables at 3, 6, and 12 months.
WHAT YOU NEED TO SUCCEED IN THIS ROLE
5+ years in data engineering with end-to-end ownership of production pipelines.
Hands-on with Snowflake, PostgreSQL CDC (Debezium preferred), and Kafka.
Solid AWS experience — S3, RDS, Aurora, and cloud data infrastructure.
Data quality engineering mindset: monitoring, reconciliation, lineage.
Comfortable defining data contracts and driving requirements with backend engineering teams.
Technical leadership experience: project ownership, cross-team alignment, delivery under constraints.
Kubernetes, Airflow/Prefect/Dagster, and dbt are strong pluses.
Payments domain knowledge — settlement, transaction lifecycle, processor integrations — is a strong plus.
Familiarity with gRPC, RabbitMQ, and reading Go/Python service code is expected.
