The three stories logged on this beat share a single thread: investors in late 2026 are not funding end-user AI products, but the unglamorous infrastructure those products and the wider economy depend on. Nodal Networks and Unsiloed each raised seed rounds to build plumbing - one for container terminals, one for documents feeding AI agents - while a Crunchbase News analysis of the blue economy shows ports and maritime logistics becoming testing grounds for similarly infrastructural technology. The pattern is a shift in where seed capital is going: toward bottleneck-clearing systems, not consumer-facing applications.
Infrastructure, Not Applications
Nodal Networks, as SiliconANGLE reported, raised $11.3 million in seed funding for artificial intelligence decision networks that manage shipping container terminals, aimed at helping port and rail hubs cut congestion and move more cargo. The company's pitch rests on the scale of the problem it targets: freight networks move more than a billion shipping containers per year. Unsiloed, also per SiliconANGLE, raised $7.5 million in seed funding to turn PDFs and other complex documents into structured, readable data for artificial intelligence agents, in a round led by Felicis with participation from Y Combinator, Orange Collective, Uncorrelated Ventures, Aarthi Ramamurthy and Beth Turner.
Neither company sells a chatbot or a consumer app. One sells throughput at ports; the other sells clean inputs for software agents that otherwise choke on scanned paperwork. Both are enabling layers. The common bet is that the returns in this cycle accrue to whoever removes friction from the systems everyone else must use.
The Billions Behind the Bottlenecks
The scale figures in the source material explain why investors are willing to fund seed-stage infrastructure. A freight system moving more than a billion containers a year, as SiliconANGLE notes of Nodal's market, is large enough that marginal improvements in congestion carry real economic weight. The Crunchbase News piece puts the blue economy at $2.6 trillion, with blue tech funding nearing $10 billion through the third quarter of 2026, driven largely by major defense deals. It also identifies maritime automation, clean energy and logistics as the startup opportunities that figure is creating, with ports emerging as hubs for testing and scaling these technologies.
That is a lot of capital chasing a sector defined by physical constraints: terminals, vessels, paperwork, energy. The report's author, Alberto Onetti, chairman of Mind the Bridge, argues that realizing the potential requires coordination among port authorities, startups, investors and policymakers. That is an unusual constraint for venture capital, which prefers to fund a single company and let the market sort out the rest. Here the market is public infrastructure, and the buyers are often public or quasi-public entities.
Why Documents Are Infrastructure Too
Unsiloed's round is the clearest illustration of the thread, because its product is invisible by design. The company turns PDFs and other complex documents into structured data for AI agents, per SiliconANGLE, addressing the fact that documents arrive in a variety of formats. Every enterprise deploying agents in 2026 hits this wall: the model is capable, but the input is a scanned invoice, a faxed form or a legacy PDF. Unsiloed sells the shovel.
The investor list is instructive. Felicis led, with Y Combinator, Orange Collective, Uncorrelated Ventures and two individual investors participating. Y Combinator's presence is a reminder that this is a well-worn seed-stage archetype - the picks-and-shovels company - rather than a novel thesis. What has changed is the buyer. In 2026 the customer is not a data team doing analytics but an agentic system that needs machine-readable input to function at all.



