Exclusive: Seltz, a startup rebuilding web search for AI agents, raises $12.5 million in seed funding

Seltz, a web search startup founded by veterans from Amazon and Pinecone, raised $12.5 million in seed funding led by Speedinvest and Capital B. The company is building search infrastructure specifically designed for AI agents rather than human users, addressing a gap in how autonomous systems access and process web information.

Seltz's funding round reflects a critical emerging need in the AI agent ecosystem. As large language models increasingly power autonomous agents that need real-time web access, existing search infrastructure designed for human consumption becomes suboptimal. Traditional search engines prioritize user experience and engagement metrics; Seltz is architecting search for machine consumption, optimizing for accuracy, structured data extraction, and integration with AI reasoning pipelines. This represents a fundamental shift in how information infrastructure must evolve alongside AI capabilities.

The founding team's pedigree—drawing from Amazon's scale expertise and Pinecone's vector database specialization—signals that this isn't a marginal improvement but a ground-up rethinking of search. As AI agents become central to enterprise automation, productivity tools, and autonomous research systems, the quality and efficiency of their information retrieval becomes a critical bottleneck. Current solutions rely on API integrations with consumer search engines that weren't built for the latency, format, and scalability requirements of agent systems.

This funding validates investor confidence that AI agent infrastructure is a multi-billion-dollar market opportunity. Speedinvest and Capital B's involvement suggests conviction that search-for-agents will become as essential as vector databases were for the AI boom. The implications extend beyond Seltz—it signals that AI infrastructure layers are still being defined and competed for. Developers building agent systems will likely face an increasing array of specialized tools optimized for different aspects of autonomous reasoning, creating a new category of AI infrastructure startups that could attract significant capital.

• →Seltz is building search infrastructure specifically optimized for AI agents, not human users, filling a gap in the AI ecosystem
• →The $12.5 million seed round led by Speedinvest and Capital B validates a new category of AI infrastructure startups
• →Founding team expertise from Amazon and Pinecone indicates this is a fundamental rethink of search architecture rather than incremental improvement
• →AI agents require different search characteristics than humans—better structured data, faster APIs, and tighter integration with reasoning systems
• →This funding round suggests investor confidence that agent-specific infrastructure will become as critical as vector databases in the AI stack