Senior Software Engineer - Multi-Cloud Compute Platform

Databricks; Mountain View (CA); On-site; USD 164,000 - 205,000; Full time

Job summary

Databricks is seeking an engineer for the Serverless compute team to design and build a next-generation, highly available platform that scales across AWS, Azure, and GCP. You will tackle sub-second startup times, multi-tenant security isolation, and a unified orchestration layer across clouds.

You’ll work on a single compute stack that supports AI training workloads, stateful execution for Lakebase, and a host-level optimization of networking and orchestration, collaborating with leaders from

Qualifications

• 5+ years of production-level experience in programming languages such as Python, Java, Scala, or C++.
• 4+ years building large-scale distributed systems from scratch.
• Experience with Kubernetes, Docker, or MicroVMs and understanding cloud provider internals.
• Able to build tools/platforms at scale when existing solutions fail.

Responsibilities

• Design and build the next generation of our Serverless platform with high availability and low latency.
• Create a unified compute stack supporting heterogeneous workloads including AI training.
• Develop platform-level abstractions for multi-cloud deployment across AWS/Azure/GCP.
• Optimize APIs, networking, and container orchestration from top to bottom.
• Collaborate with experts who built systems like Borg, Kubernetes, and Uber’s infra.

Skills

Python/Java/Scala/C++

Distributed systems

Kubernetes/Docker/MicroVM

Cloud internals (AWS/Azure/GCP)

Job description

Databricks is seeking an engineer for the Serverless compute team to design and build a next-generation, highly available platform that scales across AWS, Azure, and GCP. You will tackle sub-second startup times, multi-tenant security isolation, and a unified orchestration layer across clouds.

You’ll work on a single compute stack that supports AI training workloads, stateful execution for Lakebase, and a host-level optimization of networking and orchestration, collaborating with leaders from