Data Engineer

About Ardenus

Ardenus engineers the integration layer that makes every external data source - relational, transactional, API-based, file-based, third-party SaaS - land in a typed, query-ready form on a continuous basis. The role exists because data integration is the single largest source of risk in production analytics systems, and we've decided to solve it once.

As a Data Engineer, you own the ingestion plane end-to-end. This is not a tool-configuration role and not a data-warehouse role. It is the engineer responsible for the contracts, the pipelines, and the reliability of the layer everyone else's work sits on top of.

Description

• Architect and operate the ingestion plane - connectors, change-data-capture, schema discovery, replication, and reconciliation across heterogeneous source systems.
• Design the contracts between raw source data and downstream typed entities, including upserts, deletes, late-arriving facts, and temporal versioning.
• Ship continuous pipelines that run unattended at meaningful scale, with measurable freshness SLAs and provenance guarantees.
• Own data quality across the platform - reconciliation, drift detection, schema evolution, and the alerts that fire before customers notice.
• Define the storage and partitioning strategy that supports both transactional access and retrieval workloads.
• Set the engineering standard for pipeline reliability, schema evolution, and data correctness.

Minimum Qualifications

• You should:
  • Have 7+ years building production data pipelines, with at least 3 of those years owning a multi-source ingestion layer end-to-end.
  • Have personally owned a pipeline that processed in excess of 10 TB per day, or its equivalent in event throughput, and be able to describe the specific failures you engineered around.
  • Hold deep, production-grade fluency in PostgreSQL or a comparable relational system - query planning, index design, partitioning strategy, replication mechanics.
  • Be cloud-native in a major cloud (AWS, GCP, or Azure) across object storage, managed compute, and managed databases.
  • Treat schema evolution and tenant isolation as engineering disciplines, not configuration concerns.

Preferred Qualifications

• You can:
  • Point to public technical work - a contribution to a major CDC or streaming project (Debezium, Kafka Connect, Flink, Arroyo), a published post on data architecture that's been cited, or a conference talk.
  • Show production experience with open table formats (Iceberg, Delta, Hudi) including writer-side correctness under concurrent access.
  • Reason fluently about exactly-once semantics, idempotency, and the engineering cost of each.
  • Demonstrate experience with data contracts, schema registries, or formal data versioning at organizational scale.
  • Walk through a pipeline you inherited in a broken state and turned into one nobody thinks about anymore.