Causality Link, a provider of artificial intelligence-driven investment technology, plans to build out its research tool after a $5 million Series A funding round.
CEO Pierre Haren says the firm plans to bolster its current research platform and make some key strategic hires with the capital.
“We are accelerating our detailed domain knowledge and trading services, as well as our NLP core detection system,” Haren says.
The funding round was led by Horizon, a French family office. Horizon’s director, Jean-Baptiste Duzan, is now a member of the Causality Link advisory board.
The AI-powered research platform, Causality Link Research Assistant, is designed to extract insight and forecasts on companies, industries and macroeconomic events from texts by highlighting specific cause-and-effect relationships between market indicators and company key performance metrics.
The platform uses machine learning and natural language processing (NLP) techniques to combine the insights of industry experts with changes in the markets. The information is used by quants, investors, financial analysts and researchers to inform investment decisions.
The platform processes an average of 40,000 texts per day in 18 different languages. Content aggregator Naviga provides Causality Link with these texts in real time, corresponding to roughly 4,000 newspapers, magazines, other publications and government sites that are updated daily.
The software pings the Naviga service every second to download new texts before translating them into English, filtering and analyzing them. The resulting metadata is sent to an elastic database that is the real-time source of information for all Causality Link’s services.
“We are going to spend some money on more content. Our system can digest four million texts per day. Today we ran through 40,000. So there is capacity for a lot more,” Haren says.
Haren says the quality of the data in the system is high because all the information comes from legitimate publications rather than solely from web scraping. Tweets and citizen journalism, for instance, are filtered out of the analytics unless they are significant enough to have been written about by a reputable publication.
“We don’t read tweets. We read the posts that serious people make about the tweets,” he says.
Causality Link’s software runs on AWS’s Athena, a serverless query service that allows data analysts to perform SQL queries on data in Amazon public cloud.
To read the full article, click here.