Text size

The decision to invest in a company can rely on a lot of guesswork, but Kim Polese, co-founder and chairman of CrowdSmart, is using artificial intelligence to turn qualitative information into quantitative dataand reduce bias along the way.

When were talking about using collective intelligence, augmented collective intelligence, what were really talking about is using a combination of human and machine intelligence to improve the way that diligence is done, Polese said this past Wednesday at a BarronsInvesting in Tech panel. The founder of an artificial-intelligence platform designed to predict a companys potential for success, Polese detailed how the CrowdSmart platform works, and how it could help remove bias from the diligence process.

The system draws on the insights of a group of 25 or more people, selected for their different levels of expertise, to evaluate prospective investments, explained Polese, who said her career in Silicon Valley began 30 years ago at the first artificial-intelligence company to go public.

Those people are able to access all of the full diligence materials, so that might be videos, live Q&As with the teams, all of the financials, and, ultimately, a brainstorming process is kicked off, Polese said. Participants are given prompts, like do you find this a compelling investment opportunity? and what are your top concerns? to assist in evaluating the companies.

By ranking the anonymous responses that come in, investors can start to drill down into those specific elements within this investment opportunity, Polese said.

Using natural language processing, the insights gathered are transformed into a quantitative score, which can determine the investment risk or opportunity.

While the platforms primary goal was to accurately predict investment success, one side effect has been the reduction of bias, she said. Traditionally, venture-capital funding has been very much a relationship-driven, network-driven business that can leave behind underrepresented founders without connections in the industry, Polese said.

When Polese first used the platform to pick investments about four years ago, she said 42% of the highest-scoring companies were founded or led by women. That result was not something we set out to achieve as a goal, [but] a side effect of reducing ingrained bias, which is an important element of this approach, she said.

The diligence process takes place over the course of a couple of weeks and is designed from the ground up to be virtual, remote, said Polese. It can be applied to companies in different stages, from start-ups to public offerings.

By scaling diligence this way, you dont have this tiny little funnel that only a few deals can get through, Polese said. Youd have a much wider funnel that then you can evaluate with more predictive accuracy.

Email: editors@barrons.com

Original post:

How Artificial Intelligence Could Lead to Better Investment Decisions - Barron's