AI search startup Keenable has exited stealth mode with $26 million in seed funding led by Accel to build web search infrastructure tailored for AI agents. Co-founded by former Yandex search head Andrey Styskin and Amazon AGI engineer Matthias Petri, Keenable operates an independent index of over 100 billion documents. The startup aims to address the high costs and limitations of traditional search APIs by offering low-latency, model-ready inputs and a proprietary Web Query Language.
Keenable $26M seed funding
- ▪AI search startup Keenable raised $26 million in seed funding led by venture capital firm Accel to build web search infrastructure for AI agents.
- ▪Keenable plans to use its $26 million seed funding to double its 15-person engineering and go-to-market team and support its web-crawling and indexing infrastructure.
- ▪Keenable's $26 million seed funding round included participation from Conviction Partners and angel investors from companies including Amazon, Google, ClickHouse, Databricks, Snowflake, and SpaceXAI.
AI agent search infrastructure
- ▪Keenable is developing Web Query Language, a tool designed to help artificial intelligence models search, connect, and synthesize information from thousands of web sources.
- ▪Keenable has signed commercial contracts with several artificial intelligence labs and partnered with voice artificial intelligence company Gradium to support live information retrieval.
- ▪Keenable supports point-in-time historical queries, allowing artificial intelligence models to search the internet as it existed at a specific point in the past.
- ▪Keenable offers its search application programming interface at a price of $1 per 1,000 requests, targeting frontier-scale users with low-latency, model-ready inputs.
100 billion document index
- ▪Keenable operates an independent web search index spanning more than 100 billion documents, optimized for low-latency, high-frequency queries conducted by artificial intelligence agents.
- ▪Keenable's independent web search index is designed to serve specific use cases including market mapping, pricing monitoring, and lead enrichment.
Yandex Amazon founder experience
- ▪Keenable was co-founded by Chief Executive Andrey Styskin, former head of search at Yandex, and Chief Scientist Matthias Petri, a former Amazon AGI search engineer.
- ▪Keenable co-founder Andrey Styskin previously built a 200-billion-document index at Yandex and worked with Matthias Petri on web search infrastructure for Amazon Alexa.
Traditional search API limitations
- ▪Keenable Chief Executive Andrey Styskin stated that enterprise-scale search systems can become prohibitively expensive at web scale if index structures are not fine-tuned for specific tasks.
- ▪Traditional search application programming interfaces are unsuitable for artificial intelligence agents because they were designed for occasional human queries rather than high-frequency, repetitive tasks.
- ▪Traditional search queries can quickly become expensive for artificial intelligence agents, which encourages models to rely on static training data instead of retrieving current web information.
Agentic search market opportunity
- ▪Keenable competes in the agentic search market against rival platforms including Tavily, Exa Labs, Perplexity AI, and Brave.
- ▪According to Accel partner Zhenya Loginov, Google and Microsoft are retiring or shutting down their existing search application programming interfaces to avoid cannibalization.
Story comments
Loading comments…