Conference Presentation, Panel, Fireside Chat
OVH Cloud, ClickHouse, Neon, Aiven, Airbyte & Commit: Open Software Infrastructure 2.0
- AI agents are expected to become the primary data consumers, shifting usage from human-run dashboards to direct agent access within their context.
- Industry infrastructure is predicted to transition from compute-centric to data-centric models, requiring providers to scale networks and interconnections specifically for agentic AI workloads.
- Open source software deployment at scale is forecasted to continue growing, making it increasingly difficult for smaller companies to justify managing infrastructure over core product development.
- AI models are anticipated to become a local commodity runnable on consumer devices within five to ten years, rendering efficiency concerns less critical as the technology becomes ubiquitous.
- The population capable of building applications is projected to expand from approximately 30 million to hundreds of millions, necessitating a strategic shift from professional developers to "vibe coding" users.
- Reduced friction and costs in data R&D are expected to accelerate corporate reaction times from five days to one minute, fundamentally altering operational models.
- Current vertical integration trends among startups are predicted to yield stable, common infrastructure patterns over the next five years.
- While AI application builders may evolve into platforms for hosting infrastructure, it remains premature to determine if they will become the primary software distribution channel.
- Future product-market fits for many innovations are likely to be discovered by individuals lacking formal training in software development or infrastructure management.
- The current high cost and complexity of AI are expected to evolve into a seamless, omnipresent utility where users no longer need to manage model efficiency.
- Disruption in streaming systems is predicted to occur if new implementations adopt existing protocols like Kafka, though this shift is expected to be minimal.
- Software construction is expected to shift further away from hardware procurement and IT department management toward service-based, automated provisioning.