From Single-Player Chatbots to Enterprise AI Platforms
The latest funding rounds for Dust and Hark underscore a structural pivot in how enterprises are adopting AI. Instead of stitching together isolated tools, organisations are gravitating toward an enterprise AI platform model that centralises orchestration, governance and AI agent deployment. Dust positions itself as an operating system for AI agents, while Hark is building a universal interface that blends models, software and AI-native hardware. Both approaches reflect investor conviction that the next wave of value will come from platforms that sit above individual models, connect deeply into enterprise systems and support persistent, context-aware interactions. This shift also reframes AI infrastructure investment: capital is increasingly flowing into companies that abstract away model complexity and provide shared environments where humans and agents collaborate, rather than into narrow, single-purpose point solutions. The combined momentum around Dust and Hark suggests this platform-centric architecture is becoming the new default for enterprise AI deployments.
Dust’s Multiplayer Operating System and Explosive Agent Adoption
Dust has raised USD 40 million (approx. RM184 million) in Series B funding led by Abstract and Sequoia to scale what it calls a multiplayer operating system for enterprise AI. The platform enables companies to deploy, orchestrate and govern fleets of specialised AI agents that share context with human teammates. More than 3,000 organisations now use Dust, with over 300,000 agents deployed across its environment. Critically, Dust reports 70% weekly active usage and zero churn in 2025, indicating that AI is embedded into daily workflows rather than being a one-off experiment. The product centres on a shared collaboration surface where humans and agents work in the same projects, conversations and notifications, backed by a cloud compute layer for complex tasks. An intelligence layer links to over 100 data sources and tools, while enterprise governance features deliver granular permissions, auditing and cost controls—exactly the capabilities buyers expect from a modern enterprise AI platform.
Hark’s Universal Interface Strategy and AI-Native Hardware Bet
Hark has secured more than USD 700 million (approx. RM3.22 billion) in an oversubscribed Series A round at a USD 6 billion (approx. RM27.6 billion) valuation, reflecting extraordinary confidence in its universal interface vision. Rather than focusing on a single layer of the stack, Hark is building a platform that fuses AI models, software and custom hardware designed for personalised, persistent human–machine interaction. The company, now roughly 70 employees strong, has added a new NVIDIA B200 data centre to train its own models and plans to release its first agentic, multimodal systems later this summer. Hark’s ambition is to create assistants that remember who users are, manage their digital lives and operate across existing products and services with near human-level intuition. This hardware–software co-design approach positions Hark not just as another model provider but as an end-to-end AI infrastructure layer that could redefine how users interface with their entire digital environment.
Why Investors Are Backing Platforms Over Point Solutions
Taken together, Dust’s and Hark’s rounds represent over USD 740 million (approx. RM3.4 billion) in fresh capital directed squarely at the enterprise AI infrastructure layer. Investors now appear less interested in single-purpose chatbots or niche automation tools and more focused on platforms that can support broad AI agent deployment, governance and integration across an organisation’s tech stack. Dust’s metrics—70% weekly active usage and zero churn—are especially telling: they show that once a robust platform is in place, AI usage compounds across teams and use cases. Hark’s universal interface strategy points in the same direction, suggesting that AI will be experienced as a continuous layer over work and life, not as isolated apps. For investors, this platform-first orientation promises durable network effects, higher switching costs and recurring revenue, making enterprise AI platforms one of the most compelling bets in today’s AI infrastructure investment landscape.