(Bloomberg) -- Bridgewise, a startup that uses artificial intelligence to provide investment research for global securities, has raised $21 million in funding, as the finance industry expands its adoption of the technology.

The Tel-Aviv based startup uses machine-learning algorithms trained on historical data to provide investment analysis to brokerages, wealth advisors and exchanges such as Nasdaq, the London Stock Exchange and the Tel Aviv Stock Exchange. It’s also created a robo-advisor named Bridget that it claims can deliver on-demand research of any financial security, personalized for an individual or family office portfolio. The service works like ChatGPT, where users can ask almost any question, but it offers the kind of financial guidance OpenAI’s chatbot often steers clear of. 

Bridgewise, founded in 2019, will use the new money to accelerate its global expansion and develop new AI user tools. It says it now has 50 clients and is used in 35 trading platforms.

“Millions look at our AI-generated reports daily,” Gaby Diamant, the firm’s 35-year-old chief executive officer, said on a video call from Tel Aviv. “We help investment advisers focus on fundamentals, whether they speak a certain language or not, whether they understand all the jargon or not.”

Bridgewise provides financial reports in 22 languages across 15 countries such as Australia, Japan, Singapore and the US. It covers 36,000 equities and 14,500 exchange-traded or mutual funds, and says its research is updated daily with the latest information from news reports and social media.  

The series A funding was led by SIX Group, an infrastructure operator in the Swiss and Spanish financial centers, and included Group 11 and L4 Venture Builder. The backers are finance companies that will become strategic partners in certain regions to help access those markets, Bridgewise said. It has raised a total of $35 million so far.

Bridgewise also offers a kind of generative AI that it calls a micro language model, or MLM, that delivers analysis in the reader’s preferred language in a conversational chat. While a large language model such as OpenAI’s GPT-4 uses more than a trillion parameters, its MLM uses a much smaller data set to save time and money, as well as to concentrate its expertise. Diamant said Bridgewise is building large language models (LLMs) for specific geographies and customers.

©2024 Bloomberg L.P.