Lead Data Architect

Role Summary

Own the target-state data architecture on Google Spanner, ensuring the new data model satisfies functional correctness, performance SLOs, and transactional semantics required by 1000+ applications—before large-scale app rewrites occur.

This role is not a schema converter. It is a distributed systems architect role.

Key Responsibilities

Data Architecture & Modeling

• Define the Logical Canonical Data Model (LCDM) independent of physical storage.
• Design Spanner-native physical schemas:
• Primary key strategy
• Row key distribution (hotspot avoidance)
• Interleaving decisions
• Secondary index strategy
• Decide where denormalization, projections, or materialized views are required.

Transaction & Consistency Design

• Define transaction boundaries under Spanner constraints:
• Single-partition vs multi-partition transactions
• Atomicity and consistency trade-offs
• Classify workloads by strong consistency vs bounded staleness tolerance.
• Define idempotency and retry-safe patterns.

Compatibility & Contract Validation

• Work with QA and platform teams to validate:
• CRUD contracts
• Query semantics
• Transaction invariants
• Review and sign off on schema compatibility scorecards.

Performance & Scalability

• Participate in workload replay analysis.
• Diagnose and resolve:
• Hot key risks
• Transaction conflicts
• Index amplification
• Size Spanner instances and estimate cost envelopes.

Leadership

• Act as final authority for Spanner data design decisions.
• Mentor domain architects and engineers.
• Provide executive-level architectural rationale.

Required Skills & Experience

• 10+ years in large-scale data architecture
• Deep expertise in Google Spanner internals:
• TrueTime
• Distributed transactions
• Synchronous replication
• Experience designing petabyte-scale, multi-region OLTP systems
• Strong understanding of:
• Distributed consistency models
• Failure modes in globally distributed databases
• Prior Oracle → distributed DB migration experience strongly preferred

Key Deliverables

• Canonical logical data model
• Spanner physical schemas & index strategies
• Transaction design guidelines
• Performance readiness assessments
• Migration readiness sign-off per domain