Keenable, a search infrastructure startup led by former Yandex and Amazon engineers, has secured $26 million in seed funding to build a web index optimised for AI agents rather than human users. The company's core insight is that AI systems require fundamentally different search infrastructure than Google's human-facing model—whilst people need concise results, AI agents can process and synthesise information from hundreds of documents simultaneously, creating what founder Andrey Styskin calls "a new flywheel" distinct from traditional search optimisation. Keenable has already indexed over 100 billion documents and deployed its API across multiple AI labs and inference providers, with an upcoming product called WebQueryLanguage designed to help AI systems answer questions by combining information across disparate sources. The timing reflects a genuine infrastructure gap: Google and Microsoft have restricted their search APIs to prevent cannibalisation of their own AI products, leaving AI developers with limited options for web-scale retrieval at production scale.
For CX teams, this development signals a critical shift in how AI agents will access and ground their responses in real-world information. As organisations deploy agentic systems across support, sales, and customer success workflows, the quality of those agents' knowledge retrieval directly impacts resolution accuracy and customer trust. The question becomes whether your current AI infrastructure—whether Zendesk AI Agents, Salesforce Agentforce, or custom implementations—will have access to reliable, cost-effective web-scale retrieval, or whether you'll remain dependent on proprietary APIs controlled by tech giants with competing interests. Keenable's emergence suggests the market recognises this as a solvable problem, but it also indicates that teams relying on closed-loop AI systems without robust external knowledge retrieval may face competitive disadvantages as agentic workflows mature.
The broader implication is that web-scale information retrieval is becoming a commodity infrastructure layer rather than a competitive moat. Styskin's argument that smaller players can outcompete Google on agentic queries through cost efficiency and purpose-built indexing mirrors how enterprise search vendors have carved out niches from Google's dominance. For CX leaders, this means the next generation of AI agents will likely depend on specialised retrieval infrastructure separate from your core platform—raising questions about integration complexity, vendor lock-in, and whether your current tech stack can accommodate multiple retrieval sources without degrading agent performance or increasing operational overhead.
Now exiting stealth mode with a $26 million seed round, Keenable has been building a vast web search index for AI agents.