Furnace emerges with $3.5 million to standardise prediction-market data

Furnace has emerged from stealth with a $3.5 million seed round to build data and reference infrastructure for institutional prediction markets. The company is targeting hedge funds, prime brokers, proprietary trading desks and developers building products on prediction-market platforms.

It describes its product as a reference layer: a system that assigns a persistent canonical identity to an underlying economic exposure, then maps the differing contracts and identifiers offered by individual venues beneath it. The approach is intended to preserve each instrument’s separate terms and resolution details while its technology determines which contracts are economically alike and which differ materially.

Furnace said its technology addresses inconsistent contract terms, settlement definitions and data feeds across competing exchanges. It is available through integrations and APIs, with data normalised for institutional risk systems.

The company covers the US and international versions of Polymarket, as well as another major prediction-market operator, and said it is adding venues. Matthew Siff said prediction markets were becoming part of the financial system without the common identifiers and reference data that institutions use in other markets.

North Island Ventures, Reciprocal Ventures, Reverie and VanEck are listed as financial backers. Commonweal, DTCC and Public.com are listed as angel investors.

The public launch fills in details that were not disclosed while the company was in stealth. Fundraise Insider, citing an SEC filing, reported that Furnace Systems Corp had raised $3.55 million from nine investors. The company was founded in 2025 by Matthew Siff, Andrew Siff and Gabriel Rizk, and is based in New York.

The launch comes as prediction-market operators and service providers seek more institutional participation in a fragmented market. Fortune reported that Kalshi’s institutional trading volume rose 800% over the six months before May, while recent block trades on Kalshi and Polymarket involved trading on the outcome of a California carbon-allowance auction and hedging exposure to GPU-rental prices, respectively. Furnace’s focus is on reference data for institutional systems.