Staff Engineer (one Level Above Senior - Backend) - App Experiences, Ecosystem

Canva Canva · Enterprise · Sydney, Australia · Information Technology

Staff Engineer role focused on building backend services for Canva's App Experiences, including policy evaluation, search, recommendations, and governance. The role involves designing and operating high-throughput services, working with databases like DynamoDB and Qdrant, and shifting towards a platform model for app discovery and management.

What you'd actually do

  1. Design and build the App Access Policy service — the new backend that governs how organisations and brands control app access across Canva, using a DynamoDB + SpiceDB architecture to handle high-throughput policy evaluation at scale.
  2. Evolve the App Search and App Installation services — our core discovery and installation infrastructure — improving data models, payload efficiency, and reliability to support faster experimentation across all App surfaces in Canva.
  3. Drive the shift from a front-end-heavy architecture to a platform model: move ranking, recommendations, filtering, and governance logic from the front-end into well-defined backend services that any surface team can consume with confidence.
  4. Contribute to semantic search and recommendations by working with ElasticSearch and vector databases (Qdrant) to deliver more personalised, context-aware App discovery — helping us move from fetching all apps and filtering them client-side to intelligent, intent-aware recommendations.
  5. Partner with front-end engineers and product teams to design clean, minimal data interfaces that reduce payload sizes, improve load times, and unblock teams building App surfaces across Canva — including in the logged-out marketplace.

Skills

Required

  • strong backend engineer
  • deep Java experience
  • designing and operating high-throughput services in production
  • reliability
  • observability
  • DynamoDB
  • access patterns
  • model data to avoid hot partitions
  • shipped production systems that depend on it
  • access control systems or authorisation infrastructure
  • SpiceDB, Zanzibar-style models, or similar policy evaluation systems
  • tradeoffs between relational and graph-based approaches to permission modelling
  • search and data infrastructure
  • ElasticSearch in production
  • vector databases (Qdrant, Pinecone, or similar) for semantic search or recommendations
  • think in platforms, not features
  • building shared primitives, clean service boundaries, and data contracts
  • reducing complexity
  • migrated legacy systems
  • cleaned up technical debt
  • architectural improvements
  • collaborative, cross-functional team
  • contributing to product and design conversations
  • writing docs that build alignment
  • mentoring less experienced engineers

Nice to have

  • co-working spaces
  • flexible leave options

What the JD emphasized

  • high-throughput policy evaluation at scale
  • high-throughput services in production
  • reliability, observability, and getting things right at scale
  • access control systems or authorisation infrastructure
  • policy evaluation systems
  • search and data infrastructure
  • vector databases
  • semantic search or recommendations
  • building shared primitives, clean service boundaries, and data contracts
  • reducing complexity
  • migrated legacy systems
  • cleaned up technical debt
  • architectural improvements that compounded over time
  • building a Marketplace Platform
  • shared services and data layer
  • foundational, high-leverage engineering work
  • shape that platform from the ground up
  • making technical decisions that compound
  • reducing complexity
  • building infrastructure that other teams across Canva will rely on