Cascade

London and New York-based AI startup Cascade raised $3.5M seed round led by Andreessen Horowitz Speedrun to predict engineering and construction projects before they become public, with additional backing from Ada Ventures, Blitzscaling Ventures, Indico Capital, shuckerVC, G2C Ventures, and Snowball VC.

Leadership read: Cascade has committed to a product bet that is structurally harder than it appears: predicting infrastructure project pipelines before public disclosure requires continuous ingestion and interpretation of heterogeneous signals, bond filings, permit databases, earnings transcripts, municipal meeting minutes, across two jurisdictions with different data regimes. That is not a search or summarisation problem; it is a structured-signal extraction and classification problem that has to be right enough, early enough, to change commercial behaviour. Surfacing $10bn in claimed opportunities means the data coverage question is already live; the funding commitment is to expand that coverage, not to prove the core thesis. The related signals in this 90-day window are broadly capital-markets events with limited thematic overlap; this is one of the few early-stage AI-for-infrastructure signals in the set. The more relevant comparable pattern sits slightly outside the window: a cluster of construction-tech and project-intelligence tools that have raised seed and Series A rounds positioning around pre-procurement intelligence, several backed by specialists in applied AI for regulated industries. The a16z Speedrun imprimatur is notable precisely because it accelerates credibility in US enterprise sales, which appears to be a deliberate geographic anchor alongside the UK base. Companies reaching this stage in AI-for-infrastructure commercial intelligence consistently surface demand for leadership across data partnerships and procurement-signal sourcing, enterprise GTM with long-cycle construction clients, and product management at the seam between signal quality and revenue attribution, the last of these being where churn risk concentrates when early claims meet client expectations at scale.

curated · 2026-07-21 · context →