Data Lake Backend Engineer - Storage & Query

Kloudfuse; Cupertino, Northern (CA, KY); Hybrid; USD 140,000 - 210,000; Full time

Job summary

Kloudfuse seeks an experienced systems engineer to own the storage and query engine of our observability data lake in Cupertino. You will shape ingestion pipelines, segment layouts, indexing, and tiering across hot and cold storage to keep 300 TB/day responsive.

You’ll push the distributed planning and execution, implement encoding schemes, and design rollups that scale millions of series while maintaining correctness and observability. On-call ownership is part of the design.

Qualifications

• Deep hands-on experience with distributed data systems such as DBMS, stream processors, or time-series stores.
• Fluency in fundamentals: consensus, replication, partitioning, consistency models and failure semantics.
• Performance instincts grounded in measurement and profiling to minimize bytes and cache misses.
• Comfort with Go or Java at production depth and willingness to work in either.
• On-call ownership of systems is expected as part of the design feedback.

Responsibilities

• Own the storage layer of the observability data lake: ingestion pipelines, segment layout, indexing and tiering across hot and cold storage.
• Push the query engine: distributed planning and execution, predicate pushdown, vectorised scans and cost modeling.
• Design encodings, sketches and rollups to keep millions of series affordable at scale.
• Build multi-tenant isolation with quotas, admission control and backpressure that fail predictably.
• Make correctness observable: consistency guarantees, replay and repair paths, and tests before customer exposure.

Skills

Distributed systems

Go

Java

Production on-call

Observability

Tools

Apache Pinot

Druid

ClickHouse

Kafka

Flink

Job description

Kloudfuse seeks an experienced systems engineer to own the storage and query engine of our observability data lake in Cupertino. You will shape ingestion pipelines, segment layouts, indexing, and tiering across hot and cold storage to keep 300 TB/day responsive.

You’ll push the distributed planning and execution, implement encoding schemes, and design rollups that scale millions of series while maintaining correctness and observability. On-call ownership is part of the design.