Keenable, a startup backed by Accel, has emerged from stealth with a $26 million seed round. The company is developing a comprehensive web search index specifically designed for AI agents, aiming to improve how AI systems access and utilize web information.
SynthePulse Insight · AI deep reading
Version 1 · 1 source
When search engines were designed for humans, AI agents need new infrastructure. Keenable, with $26 million in seed funding and an index of over 100 billion documents, aims to become the 'next Google for AI agents.'
Keenable recently emerged from stealth, announcing $26 million in seed funding led by Accel, with participation from Conviction Partners and select angel investors. The company claims to have built a web search index containing over 100 billion documents, with its API already in production use at multiple AI labs and inference providers, covering both training and inference phases. The company has not disclosed its customer list, but recently partnered with voice AI company Gradium to support real-time information retrieval.
Founder Andrey Styskin previously led the search, AI, and cloud divisions at Russian search giant Yandex, and co-founder Matthias Petri is a German AI scientist. Styskin has 20 years of experience building search at Yandex and Amazon. He believes AI chatbots perform better when backed by source documents, creating a new flywheel different from what Google learned from human behavior.
Unlike enterprise search solutions that can be costly and collapse at scale, Keenable emphasizes fine-tuning index structures for specific tasks to quickly narrow the search space, reducing the cost of serving the entire internet. Styskin says that without task-specific index tuning, the cost of serving the entire internet would be enormous due to data volume.
Accel partner Zhenya Loginov notes that AI companies have few options for web-scale search infrastructure, as Google and Microsoft are shutting down existing search APIs to avoid self-cannibalization, instead adopting bundling and selective partnerships. Styskin believes these decisions confirm his judgment, formed while working with Petri at Amazon, that there is an opportunity to build search infrastructure for AI. He has recruited several former colleagues to join the startup through his network.
Keenable's dream is to become the 'next Google for AI agents,' but it faces competitors like Brave and Exa, and Google is also overhauling its search experience for the AI era. Styskin acknowledges that while getting users to leave Google is 'extremely difficult,' the innovator's dilemma means Google may be 'beatable' on agentic queries, and smaller companies can offer more cost-effective solutions.
The cost of building a large search index is real, with Styskin describing it as 'painfully expensive,' but the company is working to control costs and maintain momentum. The team currently has 15 engineering staff spread across the US and Europe, and plans to double headcount by year-end using the new funds to build go-to-market capabilities.
This article's information primarily comes from TechCrunch's report, which includes company statements and investor perspectives. Some data (such as index size and API usage) has not been independently verified and should be considered company claims.
Keenable's funding and product progress indicate that rebuilding search infrastructure for AI agents is becoming an investment hotspot, but the company faces the dual challenges of high costs and competition from giants.
Primary report
Primary source