A New Search Index Built for AI Agents

Keenable, a startup that emerged from stealth with $26 million in seed funding, is building a web search index tailored for AI agents. The company aims to address a growing need for infrastructure that lets AI systems retrieve and process vast amounts of online information efficiently.

Founders with deep search pedigree

Keenable was founded by Andrey Styskin, who previously led Russian search giant Yandex's search, AI, and cloud division, and German AI scientist Matthias Petri. Styskin brings two decades of experience building search systems at Yandex and Amazon.

Accel led the seed round, with participation from Conviction Partners and unnamed business angels. Accel partner Zhenya Loginov led the investment for the firm.

Massive index for AI consumption

Keenable claims to have built a web search index of more than 100 billion documents. The startup says its API is already in production at several AI labs and inference providers, used during both training and runtime, though it declined to name customers.

Keenable recently struck a partnership with voice AI company Gradium to support live information retrieval, a move that underscores its focus on real-time AI applications.

Challenging the economics of web-scale search

Styskin argues that enterprise search solutions break down at web scale due to cost. "If you do not fine-tune your index structures for a specific task, the cost of serving and scanning the whole internet is enormous because of the volume. That’s why you need to innovate on how you can narrow the search space based on your query very fast. This is what we are bringing to the table," he said.

He emphasized that Keenable's approach differs by optimizing index structures for specific tasks, reducing the computational load of scanning billions of documents.

Big tech's API retreat

According to Accel partner Zhenya Loginov, AI players have few options for web-scale search infrastructure. Google and Microsoft have taken steps to shut down their existing search APIs, instead opting for a more bundled approach and being selective about partners.

Styskin saw this opportunity while working with Petri at Amazon on web search infrastructure for AI applications such as Alexa. He noted Cloudflare data showing that AI crawlers were responsible for a growing share of search volume, which reinforced his conviction.

Building proprietary retrieval

Keenable is developing proprietary retrieval capabilities, including an upcoming product called WebQueryLanguage. This tool would help AI systems answer questions by combining information from various web sources, even when none contain the full answer.

Styskin tapped into his network to hire former colleagues for the startup, building a team of 15 engineering staff across the U.S. and Europe. The company plans to use the fresh cash to double its headcount by the end of the year to build its go-to-market motion.

The cost of the index

Building a giant search index is expensive. "Don’t ask — it is painfully expensive," Styskin said, but added that the startup is doing its best to keep costs in check and pace itself.

Competition and the future of search

Keenable faces competition from players like Brave and Exa, and Google is overhauling its search experience for the AI era. Styskin acknowledged it is "extremely hard" to convince people to move away from Google for traditional search, but he argued that the innovators’ dilemma makes the U.S. giant potentially "beatable" on agentic queries.

He believes a smaller company like Keenable can innovate and offer a more cost-efficient solution to AI companies, a key advantage as demand grows.

Why it matters

The shift toward AI-native search infrastructure could reshape how AI companies access the web. If Keenable succeeds, it may offer a more cost-efficient alternative to relying on big tech's bundled search products, potentially changing the economics for AI labs and inference providers. The race to dominate agentic search is still early, but the stakes are clear: whoever controls the index may control what AI agents see and do.