A Quieter Challenger in the Enterprise AI Race
Glean built its reputation as the go-to enterprise search and AI platform for large organizations – the product that promised to connect every tool, every document, and every data silo into one intelligent layer. For a while, that promise was enough. Glean raised at a valuation north of $4 billion and locked in contracts with some of the most recognized names in enterprise software. But a Paris-based startup called Dust is now making a case that Glean’s model has a ceiling, and that ceiling is showing cracks.
Dust takes a different architectural bet. Rather than positioning itself as a universal search engine layered over existing tools, it builds what it calls “AI agents” – configurable, context-aware assistants that can be deployed for specific workflows, teams, or business functions. The distinction sounds subtle but it changes the economics and flexibility of deployment in ways that are starting to matter to mid-market and enterprise buyers who found Glean’s approach too rigid or too expensive to justify at scale.
The competition is real, and the timing is deliberate.
Where Glean Left Room for a Challenger
Glean’s core value proposition is organizational memory – you ask it something, it searches across your connected apps and surfaces an answer. That worked well enough when enterprise AI was still a novelty and IT teams were impressed by the idea of a unified search layer. But as AI tools matured and employees started expecting more than retrieval – they wanted action, automation, and workflows that didn’t require a human in the loop – Glean’s architecture began to feel like a first-generation answer to a second-generation problem.
Dust addresses that gap by treating agents as the primary product. Each agent can be scoped to a team’s actual data sources, tuned for a specific task, and connected to external APIs. A customer support team’s agent is not the same as a sales team’s agent, and that modularity is by design. The configuration happens at a level most enterprise tools don’t allow – giving IT and operations teams direct control without requiring a professional services engagement every time something needs to change. That kind of flexibility is becoming a real purchasing criterion, not just a feature checkbox.
Glean is not standing still. The company has been building out its own agent capabilities, and its funding gives it considerable runway to iterate. But the window between when a challenger identifies a weakness and when the incumbent patches it is exactly where deals get won and lost. Dust has been operating in that window, and by all observable signals – hiring activity, customer references surfacing on LinkedIn, and growing presence at enterprise AI procurement conversations – it has been converting.
The Model That Makes Dust Harder to Ignore
What makes Dust a genuine competitive threat rather than just another well-funded alternative is how it handles the integration and customization layer. Enterprise AI buyers have learned, often at cost, that out-of-the-box AI tools require significant internal investment to actually work – data pipelines need cleaning, permissions need mapping, and the AI needs context that isn’t stored anywhere neatly. Dust’s design acknowledges that reality rather than promising to abstract it away. Buyers who have been burned by tools that overpromised on setup simplicity are responding to that honesty.
There is also a pricing conversation happening. Glean operates on an enterprise contract model that works for organizations deploying at scale but creates friction for teams trying to run limited pilots or departmental rollouts. Dust’s structure allows for more targeted deployments, which means a procurement team can greenlight a single use case without committing to platform-wide adoption. That entry point changes who can say yes and how quickly, which has real implications for pipeline velocity at a moment when enterprise software budgets are being scrutinized more carefully than they were two years ago.
The agent-centric approach Dust uses is gaining broader traction across the enterprise AI market. Companies that built early products around retrieval and search – whether in customer support, internal tools, or knowledge management – are all contending with buyers who now expect those products to do something with the information, not just find it. Intercom’s push into AI agents is creating similar pressure in the customer support space, suggesting this is less a Dust-specific story and more a platform-level reckoning playing out across multiple enterprise software categories simultaneously.
What Comes Next in This Fight
Glean holds real structural advantages – deep integrations, an established enterprise sales motion, and enough brand recognition that it often shows up on shortlists by default. Dust has to earn every deal by outperforming, not just by showing up. But the more interesting question is whether the enterprise AI market bifurcates: one tier for platforms that serve as organization-wide infrastructure, and another for modular, agent-driven tools that teams deploy independently without waiting for centralized IT approval. If that split happens, Glean and Dust are not necessarily competing for the same buyer anymore – and Dust may be building the right product for the tier that grows faster.
