Private markets are relative newcomers to digital transformation, despite being rich in data drawn from the corners of the economy that the more technologically advanced public markets can’t reach.
The founders of young start-up BlueFlame AI recognised this gap in the financial services market and have created a generative artificial intelligence (GenAI) platform that enables participants to mine the troves of data they have accumulated over decades and give it structure so it can be integrated into wider data processes.
BlueFlame AI’s Nexus lets firms unlock information stored in board decks, fund documents and other unstructured sources, then parse it into a form so that it can be ingested into their systems, tabulated and queried. It uses a variety of large language models (LLMs) to establish the best answers to queries and makes the outputs available in multiple ways, including on spreadsheets and through APIs.
New Tool
The platform builds on the two-year old company’s integration platform and AI tools, said chief executive Raj Bakhru.
“The search that we had was much more like a Google search, but it wasn’t really great for cross-sectional or time-series based questions,” Bakhru told Data Management Insight. “So we built this new tool that will automatically parse the content out of those documents and turn it into a structured table of data so that you could ask those types of questions against it.”
Alternative markets, which include private markets, are growing in importance to the global economy. In 2022, they accounted for about a fifth of global assets under management and half of all revenues generated globally. Private markets are expected to manage as much as US$15 trillion in assets by next year.
They are becoming a larger part of the institutional investment portfolio, too, as companies diversify to protect against volatility in conventional markets. About a third of all institutional money is tied up in such markets.
The difficulty of assessing private market portfolio companies, however, is finding and extracting their data. The companies are usually under less rigorous reporting regimes than listed firms and funds’ tech setups tend to be less sophisticated and more fragmented.
“All of that data is incredibly valuable and really hard to get answers against, because it’s such a mass of data,” said Bakhru. “Some of our clients have five terabytes of data sitting on their drives that they’ve never been able to harness before.
“Now, with generative AI, this extractive AI technology that we built out, they have the ability to take that five terabytes and turn it into a really, really compelling and valuable private research database.”
Bakhru said that being able to command data in this way gives investment managers insights they wouldn’t have been able to achieve using standard methods. It also brings savings because some of the processes Nexus enables has often been outsourced to third parties.
Founding Team
BlueFlame AI began building its platform about 20 months ago and it launched a year ago. The company already counts investment management firms including Lightyear Capital and Charlesbank among its clients and has a workforce of 25 based in New York.
Bakhru founded BlueFlame AI with colleagues he’d met in a career that took in Goldman Sachs and in 2014 saw him create a cyber security company that was eventually bought by buy-side compliance consulting firm ACA.
“We all come from the same space, and have deep DNA in the firm around compliance, cyber privacy, which is obviously a very big headwind on AI adoption,” he said.
The team of founders self-funded the project before receiving a US$5 million series A injection earlier this year.
While most of BlueFlame AI’s clients are in the alternatives space, Bakhru said about a quarter of them are hedge funds that focus on listed firms and that the company has also begun working with investment banks and sell-side operators.
Multiple Sources
BlueFlame AI’s Nexus supports and integrates clients’ own unstructured data, their structured enterprise data and market data feeds to give transparency across the assets they manage. That can also be combined with publicly available data such as regulatory filings and the major search engines’ news feeds.
Bakhru said that BlueFlame AI’s use of AI gives smaller firms the technological edge that only larger firms have traditionally been able to achieve with their larger budgets.
“Through automation, we’ve brought down the cost of – and the barrier to entry on – doing that, such that you can structure all of your unstructured content very quickly and cost effectively,” he said.
Subscribe to our newsletter
