Staff Software Engineer, Lakeflow Pipelines DR

RDQ427R70

Our mission at Databricks is to radically simplify the data lifecycle with a unified Lakehouse platform for data engineering, analytics, and AI. The Lakehouse addresses major challenges in enterprise data platforms, including reliability, data staleness, operational complexity, total cost of ownership, and data lock-in. Learn more about the Lakehouse architecture.

Lakeflow is a critical part of this vision, helping customers build and operate streaming and batch ETL  pipelines that power business-critical data products. As these workloads become increasingly mission-critical, customers need them to continue operating through infrastructure failures and cloud-region outages.

As a Staff Engineer on the part of the Lakeflow Disaster Recovery team, you will design and implement distributed systems that replicate and recover pipelines across regions. A pipeline is more than source code and output tables: it includes streaming checkpoints, source offsets, stateful operator state, table versions, transaction metadata, schedules, and dependencies across a dataflow graph. You will solve challenging problems involving consistency, idempotency, causal ordering, failover, failback, and safe recovery without silent data loss or duplication.

You will work in one or more of the following areas:

• Cross-region replication and recovery for Lakeflow pipelines, streaming tables, and materialized views
• Distributed consistency across pipeline dependencies, table versions, and transaction logs
• Failover and failback workflows with conservative correctness guardrails
• Deep clone, metadata reconciliation, observability, and failure-injection testing
• High-fidelity recovery simulations, game-day testing, and formal reasoning about failure modes

What we look for:

• A passion for distributed systems, databases, storage systems, streaming systems, or reliability engineering
• Strong software engineering skills in Java, Scala, C++, Go, Python, or a similar production language
• Understanding of consistency, transactions, idempotency, replication, checkpointing, and data lineage
• Ability to define and work toward a multi-year technical vision with incremental, production-quality deliverables
• 8+ years of experience working on related systems preferred
• Optional: PhD or advanced research experience in databases, distributed systems, or storage

This role offers the opportunity to build the foundations of resilient Lakehouse computing and help customers keep their data pipelines running when failure matters most.