In this panel discussion on cloud operating models and innovation, Brendan Humphreys, CTO of Canva, and Nam Je Cho, Director of Solutions Architecture at AWS, explore the evolving dynamics of scale, AI, and adaptability in cloud-first organisations. Brendan, who joined Canva in 2014 as an engineer and now leads over 2,300 engineers, emphasises the company’s founding ambition to “empower the world to design”. Canva’s cloud-native infrastructure was intentionally architected for scale from day one, beginning with Heroku atop AWS, then evolving into microservices. He notes the pragmatic decision to start monolithically with clear separation of concerns, allowing for later scalability without sacrificing speed to market. This approach aligns with ADAPT’s insights that 25% of organisations are now repatriating workloads from public cloud to hybrid environments to improve cost, compliance, and control.

Nam Je Cho highlights how Australian organisations are early and bold adopters of cloud innovation. He describes AWS’s approach through the “well-architected framework”, which focuses on five pillars: security, scalability, resilience, cost, and operational excellence. With these principles, companies can better navigate the increasingly complex cloud environment. This reflects ADAPT’s data that findings that multi-cloud complexity is rising, and that unified management remains a key challenge. On data sovereignty, both Brendan and Nam agree, the challenge is now largely mitigated, with strong infrastructure options available locally and globally to meet regulatory and customer needs.

Brendan defines Canva as an AI-first organisation. He explains its three-pronged approach to deploying AI: rapid evaluation and integration of third-party models, enabling third-party developers to contribute AI tools via Canva’s ecosystem, and building proprietary domain-specific models. Internally, Canva fosters AI adoption through tool-agnostic experimentation, human-in-the-loop processes, and shared learning across teams. Nam Je Cho reinforces the need for continual reskilling and cultural readiness for experimentation, especially as generative AI enables previously expensive or risky legacy code transformations. Brendan highlights the importance of honest communication and psychological safety as organisations adapt to rapid technological shifts, underpinned by a platform-first mindset that enables product teams to move with speed and confidence.

Key Takeaways:

• Cloud-native design enables agility and scale: Canva’s early decision to build on scalable cloud infrastructure, starting with a monolithic but well-structured architecture, allowed rapid market entry and evolution into microservices without over-engineering from day one.
• AI integration must be purposeful and decentralised: Canva’s AI-first strategy blends in-house R&D, third-party model integration, and platform enablement, while encouraging cross-team experimentation and human-in-the-loop practices to drive innovation across the organisation.
• Adaptability and skills are critical in a fast-changing environment: Both speakers emphasised the need for organisations to invest in upskilling, foster experimentation, and avoid rigid tool mandates, focusing instead on business impact and enabling teams to respond quickly to change.