Staff Software Engineer – Foundational Data Systems for AI

What You’ll Build

• Global Metadata Substrate: Define and evolve the global metadata and transactional substrate that powers atomic consistency and schema evolution across exabyte-scale data systems.
• Adaptive Engines: Architect self-optimizing systems that continuously reorganize and compress data based on access patterns, achieving order-of-magnitude efficiency gains.
• Intelligent Data Layouts: Pioneer new approaches to encoding and layout that push theoretical limits of signal per byte read.
• Autonomous Compute Pipelines: Lead development of distributed compute platforms that scale predictively and maintain reliability under extreme load and failure conditions.
• Research to Production: Collaborate with Granica Research to translate advances in compression and probabilistic modeling into production-grade, industry-defining systems.
• Latency as Intelligence: Drive system-wide initiatives to minimize latency from insight to decision, enabling faster model learning and data-driven reasoning.

What You Bring

• Mastery of distributed systems: consensus, replication, consistency, and performance at scale.
• Proven track record of architecting and delivering large-scale data or compute systems with measurable 10× impact.
• Expertise with columnar formats and low-level data representation techniques.
• Deep production experience with Spark, Flink, or next-generation compute frameworks.
• Fluency in Java, Rust, Go, or C++, emphasizing simplicity, performance, and maintainability.
• Demonstrated leadership—mentoring senior engineers, influencing architecture, and scaling technical excellence.
• Systems intuition rooted in theory: compression, entropy, and information efficiency.

Bonus

• Familiarity with Iceberg, Delta Lake, or Hudi.
• Published or open-source contributions in distributed systems, compression, or data representation.
• Passion for bridging research and production to define the next frontier of efficient AI infrastructure.