Agentic infrastructure startup Seltz raises $12.5M to help AI agents search the web for answers

2026-06-24 · Source: AI – SiliconANGLE · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Robotics & Autonomous Systems, Entrepreneurship & Start-ups · Depth: Intermediate, short

Summary

Agentic infrastructure startup Seltz Inc. announced on June 24, 2026, it secured \$12.5 million in seed funding, led by Speedinvest and B Capital, to develop specialized search infrastructure for artificial intelligence agents. Founded by CEO Antonio Mallia, Seltz aims to overcome the limitations of human-centric search engines by building a full search stack, including crawlers and ranking systems, optimized for AI algorithms that generate detailed, parallel queries. This infrastructure is designed to surface structured evidence like specific text, tables, or images, rather than just links, from deep within web pages. The company, which started with a news index, claims 89% accuracy and sub-250ms response times on its Dynamic News Search benchmark. Despite facing larger competitors like Parallel Web Systems Inc. and Exa Labs Inc., Seltz plans to use the funding for engineering, hiring, and launching enterprise sales, with a goal to scale to tens of billions of documents.

Key takeaway

For Directors of AI/ML building agentic workflows, Seltz's \$12.5 million funding highlights a critical shift towards specialized search infrastructure. Your AI agents require systems designed to extract structured evidence, not just links, from the web. Evaluate your current search integration for AI agent efficiency, considering dedicated platforms that build their own full search stacks. This approach can significantly improve agent accuracy and response times, moving beyond human-centric search limitations.

Key insights

Seltz's specialized search infrastructure optimizes web data retrieval for AI agents by building a full stack for structured evidence.

Principles

• Traditional search is architected for humans, not AI agents.
• AI agents require structured evidence, not just links.
• Specialized search stacks improve AI agent performance.

Method

Seltz built its entire search stack, including crawlers, index, retrieval models, and ranking systems, to process long, detailed, parallel AI queries and extract specific page-level material.

In practice

• Develop search infrastructure optimized for machine consumption.
• Prioritize extracting structured evidence over link lists.
• Benchmark agentic search platforms with specific accuracy and latency metrics.

Topics

• AI Agents
• Agentic Search
• Search Infrastructure
• Information Retrieval
• Seed Funding
• Seltz Inc.

Best for: CTO, VP of Engineering/Data, AI Architect, Investor, Director of AI/ML, AI Product Manager

Related on AIssential

• Quantifying the Accuracy and Cost Impact of Design Decisions in Budget-Constrained Agentic LLM Search — Budgeting search depth and retrieval quality offers more accuracy gains than expanding token limits for agentic RAG.
• Exa Lands Series C Funding - StartupHub.ai — AI agents require a new search paradigm focused on real-time, long-tail context and low latency, distinct from human-optimized search.
• Parallel Web Systems hits $2B valuation five months after its last big raise — Rapid investor confidence in AI agent-tooling startups can lead to significant valuation growth.
• Introducing Web Search on Amazon Bedrock AgentCore — Amazon Bedrock AgentCore's Web Search provides agents with current, private web knowledge via a managed connector, overcoming static training data limitations.
• On Accelerating Grounded Code Development for Research — Lexical search and structured workflows can effectively ground coding agents in dynamic, domain-specific research.
• SE Radio 719: Birol Yildiz on Building an Agentic AI SRE — AI agents excel in incident response by reasoning through novel problems, surpassing rigid runbook limitations.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AI – SiliconANGLE.