Seltz Lands $12.5M Seed to Build the Search Layer for AI Agents

The long-quiet search wars are roaring back to life, this time powered by the rise of AI agents rather than human users. Seltz, a startup founded by a former Amazon artificial general intelligence (AGI) scientist, is capitalizing on this shift with a $12.5 million seed round announced on June 24, 2026. The funding, led by Speedinvest and B Capital with participation from the Italian Founders Fund, United Ventures, and Bain & Company’s Future Back Ventures, marks a pivotal bet that the next generation of search will be architected not for people skimming ten blue links, but for large language models (LLMs) and autonomous agents that demand machine-readable, citable data. Seltz is tackling a fundamental mismatch: traditional search engines return snippets and links optimized for human browsing, while AI agents require the raw, structured information buried within web pages—tables, images, and full text—to answer questions accurately. Founder and CEO Antonio Mallia, who earned a PhD in information retrieval from NYU and worked on Amazon’s AGI team and at vector-database company Pinecone, argues that 'the revolution is back again,' evoking the early 2000s when Google’s PageRank upended the search market. His thesis is that transformer models and agent workflows demand a completely reimagined search stack, one that can handle dozens or hundreds of parallel, long-form queries and return results in a format that agents can directly cite and process. This technical focus sets Seltz apart from simpler attempts to bolt AI onto existing search APIs. The market context is urgent: as companies deploy AI agents for research, customer support, and process automation, the bottleneck increasingly lies in real-time information retrieval. A chatbot that cannot pull the latest news, product prices, or regulatory updates with precision and speed loses its competitive edge. Seltz’s seed round signals that venture capital is willing to back infrastructure purpose-built for the agent economy, even as giants like Google and Microsoft pour billions into AI. Seltz’s advantage may lie in its clean-sheet architecture and Mallia’s deep domain expertise, but the company must move fast. Google is integrating generative AI into its own search products, and competition from other startups and big tech could erode any first-mover advantage. The $12.5 million infusion will likely accelerate product development and early customer acquisition, targeting companies that are building AI agents and need a search layer that works natively for machines. Looking ahead, Seltz’s success could redefine how information flows through the AI ecosystem. If it becomes the default search layer for agents, it may unlock new monetization models beyond advertising—perhaps charging per query or subscription fees for high-quality, structured data. The bet is that whoever solves agent-native search first will own a critical piece of the next internet infrastructure, much as Google’s API dominance once defined the web era. For now, Seltz is a high-risk, high-reward play in a market that is still taking shape, but its founder’s pedigree and investor backing suggest that the race to reinvent search is officially on.