Why We Invested in Brickroad: The Infrastructure for a Liquid Data Economy

We are excited to announce our investment in Brickroad, a company building the first true data multiplexer for the AI era. Think of a multiplexer as a universal adapter, instead of AI teams wiring individually to each data source they need, Brickroad acts as a single connection point that routes the right data to the right user instantly. Brickroad has developed a data procurement software that changes how AI teams find, license, and access external data. By cutting dataset acquisition times from weeks to seconds, Brickroad is positioning itself as critical infrastructure in the AI stack.

Solving One of AI's Biggest Bottlenecks

The AI industry faces a major problem: the onerous process of  acquiring training data limits the pace of model improvement. While AI models have advanced rapidly, the process of getting the data to train them still involves manual negotiations, slow legal reviews, and a lack of quality transparency. Companies face weeks-long negotiation cycles, unknown data origins, and pricing structures that make it difficult to determine fair value.

This problem is especially important for mid-sized AI companies. They do not have the internal data teams that top-tier AI labs do, but also need more data than is available for free. Brickroad addresses this by creating a demand-based exchange that allows developers to source high-quality data as easily as calling an API.

A Data Multiplexer Built for the Agent Era

At the heart of Brickroad's value is its end-to-end automation. Unlike traditional data brokers, Brickroad provides a liquid multiplexer where data procurement is instant and transparent.  The platform offers:

• Agentic Procurement: Automated agents that handle data discovery and purchasing directly from a developer's coding environment.

• Blockchain-Verified Provenance: On-chain records that ensure legal compliance and auditable data lineage. Built on the Croissant 1.1 metadata standard used by Google, NASA, and Hugging Face, Brickroad provides transparent tracking of data origins and usage rights, essential for AI teams verifying quality and regulated buyers requiring audit trails. 

• Transparent Pricing: A shift away from private, slow negotiations toward clear, consumption-based pricing models.

Brickroad is manifesting this vision through three core products:

• Atlas:  a data value estimation layer that helps AI teams quantify the expected utility of data before they spend on compute or licensing..

• Wayfinder: a data procurement tool for agents, allowing them to identify the best data sources in real time. time

• Horizon: supplier-side intelligence showing emerging searches, benchmark gaps, and buyer interest

Market Momentum and Real Traction

Brickroad has already supported transactions for top-tier AI labs, proving that even the most sophisticated builders need this infrastructure. Beyond these initial deals, the company has built a network of over 50 integrated suppliers and is rapidly expanding its pipeline. This zero-inventory model allows Brickroad to scale quickly without the capital requirements of traditional data brokers.

Both Amazon and Microsoft recently announced plans to launch in-platform data marketplaces, confirming that this infrastructure layer will become critical for the AI ecosystem.

Team

Brickroad is led by an outstanding team. Freeman Lewin (CEO) is a former blockchain attorney who brings deep expertise in data licensing. He is  joined by CTO Michael Gordon, a mathematician and repeat founder who has built production AI systems for the UK Government and Lloyd's of London, and Chief AI Officer Luis Oala, co-creator of the Croissant metadata standard used by Google, NASA, and Hugging Face. Together, Brickroad's leadership team is exceptionally capable.

Looking Ahead

As AI agents become more autonomous, they will need the instant-access rails that Brickroad is building. The shift from manual data procurement to programmatic purchasing is already beginning; companies are moving away from months-long negotiation cycles toward on-demand data access. 

We are proud to support Freeman, Michael, and Luis as they transform the data market from a collection of slow, private deals into a transparent, high-speed multiplexer.