Forge AI Raises $11M to Crack Unstructured Data for Machine Learning

Forge AI, a Cambridge, MA-based startup aiming to decode the world’s mountains of unstructured data for machine learning algorithms, says it has raised $11 million in new funding.

Jim Crowley, Jack Crowley, and Jennifer Lum founded Forge in 2017, and it’s kept relatively quiet in the time since. The company reads and digests unstructured data. Such information, which is estimated to account for 80 percent of the world’s data, includes text-heavy public Facebook and Twitter feeds, news articles, financial reports, and regulatory filings. Forge’s tools structure the data into a feed that provides its clients’ machine learning systems with a stream of actionable data that can help guide decision-making.

“You could say we are turning a text problem into a math problem,” Jim Crowley, Forge’s CEO, tells Xconomy.

Currently, the seemingly endless streams of unstructured data are generally monitored and cleaned up into a usable form by human analysts who Crowley says are “drowning with intense ‘data-wrangling’ challenges.” Forge’s aim is to free the analysts up “to model the world in new ways, be more efficient and apply their energies to the actual analysis,” he says.

The problem of unstructured data is far from a new one for organizations turning to algorithms or machine learning systems.

Over the years, several companies have attempted to tackle the problem. They include Lattice Data, which Apple (NASDAQ: [[ticker:AAPL]]) acquired in 2017; Kensho Technologies, which was bought by S&P Global (NYSE: [[ticker:SPGI]]) in March; Boston-based Indico; and Cambridge, MA-based Tamr.

Crowley claims Forge differs from the others because its output stream is structured for algorithms, rather than for a human audience. “What we are doing is feeding intelligent machines,” he says.

Forge’s technology makes use of new natural language understanding capabilities as well as a dynamic “knowledge graph” that adjusts to new information and trends, Crowley adds.

Boston-based Underscore VC led the Series A funding round. Other participating investors included Accomplice, Boston Seed Capital, Imagination Capital, and Project 11 Ventures, Forge says. The capital will boost hiring, product development, and marketing.

Crowley says Forge is currently marketing itself to two information-dependent sectors. One is financial services and investment firms—from banks to mutual funds, hedge funds and insurers. Forge says its system is up and running at The Vanguard Group, one of the world’s largest investment management companies. Employees there use Forge’s software for client experience, internal operations, and economic analytics, he says.

The company is also targeting government organizations with its products. Crowley says confidentiality agreements keep him from naming Forge’s governmental clients.

Author: Brian Dowling

Brian is a former Xconomy editor. Before joining Xconomy, he reported on Massachusetts government and politics for the Boston Herald and previously wrote as a general assignment reporter covering everything from crime and courts to electoral politics, business, and international politics. Brian earned a master’s degree in newspaper writing from the Columbia University Graduate School of Journalism and started his career at the Hartford Courant writing about manufacturing and energy. He holds a bachelor’s degree in Philosophy and Theology from Aquinas College in Grand Rapids, Michigan.