Gamalon, a startup that says it has developed artificial intelligence-like software that avoids some of the drawbacks of popular machine learning techniques, has received a big endorsement of its approach. On Tuesday, the Cambridge, MA-based company announced a $20 million Series A funding round led by Intel Capital.
Other backers of the round included .406 Ventures and Omidyar Technology Ventures, as well as earlier Gamalon investors Boston Seed Capital, Felicis Ventures, and Rivas Capital, according to a press release.
Gamalon said it has hauled in $32 million in capital to date, which includes its seed funding round and money from DARPA and other government agencies. The company was founded in 2013 by its CEO, Ben Vigoda, the co-founder and former CEO of Lyric Semiconductor, which was acquired by Analog Devices in 2011.
Gamalon developed natural language processing software that it says large enterprises are using to more easily digest text from customer communications—think survey data, phone call transcripts, and customer service messages—and generate actionable insights that might help them better serve customers and boost business. Gamalon’s early clients include business communications company Avaya.
Startups peddling A.I.-based data analytics software are a dime a dozen these days. What makes Gamalon worth following are the reputable backers it has attracted and the approach it’s taking with its products.
According to Techcrunch, Gamalon doesn’t use neural networks to crunch reams of data; instead, it feeds text into databases and builds decision trees designed to help the software learn new concepts quickly. Gamalon says the advantages of its technique are that it requires far fewer examples to train the machine learning models, and users can see how the software extracts meaning from the data, thus combating the “black box” critique that some observers have made of certain deep learning systems. When Gamalon’s system encounters data it doesn’t understand, it flags it, and users can edit the decision tree to categorize that unknown data and train the software, the company has said.
Gamalon’s “system learns orders of magnitude faster than the state-of-the-art approaches, is auditable, and is easily extendable to specific domains,” said Todor Tashev, Omidyar Technology Ventures managing partner, in a prepared statement. “Gamalon takes advancements in the space to another level.”
Still, even if Gamalon’s software works as well as advertised, good technology doesn’t automatically translate into sales. The startup has a lot of work ahead as it tries to compete in a crowded sector. Other tech companies developing A.I. tools for businesses include startups like DataRobot and Indico, and tech giants like Google, Microsoft, Amazon, and IBM.
