Emmanuel Marot, co-founder and CEO, LendingRobot. (LendingRobot photo)

LendingRobot, an automated online investment-management service that calls itself a “robo-advisor,” is launching a hedge fund that the company says is administered without human intervention, making use of Amazon Web Services’ machine learning, compute servers and blockchain technology.

Until now, the business of Seattle-based LendingRobot has been focused on helping manage investments in so-called peer-to-peer loans. In such loans, would-be borrowers too risky for banks seek loans from individuals. Companies including Lending Club and Prosper grade loan applications by risk and make them available to individual investors, most of whom spread their risk by investing small amounts in many individual loans.

The investment is attractive, as it can yield 8-10 percent interest paid monthly. But managing the choice of those multiple loans, and investing, re-investing and collecting interest, can get complex and boring, which is why LendingRobot has automated those tasks, taking a fee for its services.

Now, LendingRobot is launching an investment vehicle of its own. Called LendingRobot Series, the hedge fund is billed as an alternative to traditional fixed-income investments. It, too, invests minute amounts in hundreds of individual, person-to-person loans, but it also includes real estate loans.

The fund is managed by LendingRobot’s robo-advisor technology, which the company said scores, invests, and manages market activity without any need for human advisors. Eliminating people from the loop reduces costs and maximizes returns, the company said.

The fund will use blockchain technology to deliver a public ledger that the company described as unalterable.

“Usually a hedge fund is a black box: you get an email each month detailing how your investment is doing, but you don’t know what the investment is,” said LendingRobot CEO Emmanuel Marot in an interview. “We display every single note, even if the investment in it is tiny. You can view each payment made on it and all the vitals about each loan. Then we calculate and display a notarized hash code, and if we tamper with a single character, that’s immediately apparent.”

Machine learning, through Amazon Web Services, sifts through the many individual person-to-person loans offered to investors each day and scores each one on 40 attributes, identifying the best ones to invest in. The day’s new loans are offered by various companies at various times during each business day. When demand requires it, AWS servers scale up quickly to run thousands of attribute assessments around those times, then scale back down to a much lower baseline level, Marot said. AWS is LendingRobot’s exclusive cloud provider.

Fund assets are held in a bankruptcy-protection vehicle. The fund will charge 1 percent of assets under management and caps fund expenses at 0.59 percent.

LendingRobot says it serves 6,500 clients with more than $120 million in assets. The company was founded in 2012 by founding CEO Gilad Golan and Marot, who first met at Microsoft and shared an interest in making personal Lending Club investments. Having to spend hours in front of a computer to monitor new loan opportunities was cumbersome, so the pair created software to automate the process.

Golan left the company in October to become engineering director at Google in Seattle. The company has seven employees and has received $3.7 million in venture funding to date.

Like what you're reading? Subscribe to GeekWire's free newsletters to catch every headline

Job Listings on GeekWork

Find more jobs on GeekWork. Employers, post a job here.