Arkestro brings game theory and machine learning to procurement with $26M investment

Procurement involves buying supplies and services you need to run your business. For the most part, companies still use email and spreadsheets or legacy tools to find the best price and come to an agreement with the supplier. That’s fine as far as it goes, but in a world of growing supply chain problems and just-in-time production, it seems there should be a way to put more sophisticated technology to work on the problem.

That’s where Arkestro comes in. It’s using automation and intelligence to find the best prices automatically. Today the company announced a $26 million Series A to help keep growing the platform.

Company co-founder and CEO Edmund Zagorin says that specifically his company uses “predictive procurement orchestration,” which involves a number of different elements. “We use machine learning, game theory and behavioral science to simulate a procurement process before it begins. We suggest pricing and commercial terms to all of the parties involved. And then we help companies reach very fast agreements with their suppliers,” he explained.

Ideally, he says that the end result is that all parties involved get what they need. “Suppliers get faster purchase orders. Procurement gets faster savings, and the business ultimately gets the parts and services they need to please their customers,” Zagorin said.

The company works with a network of global suppliers and can help customers find the right ones at the best price to maintain services and supply levels.

“Predictive procurement orchestration is really about finding the right balance. So it’s not about getting too few or too many suppliers. It’s about understanding what meaningful optionality is. And so the opposite of low-intent purchasing is high-intent purchasing, and we describe that as having meaningful choices in your supplier network, and having the data to understand the business costs and benefits of each choice, and then being able to select the optimal choice of those that are available to you for each part, and really for each purchase order,” he said.

Arkestro is essentially a self-negotiating system. It’s going to continue to look for goods or services that a company needs until it finds someone who says yes to the price and amount, starting with a list of preferred providers.

Lest you think this is just about a manufacturing scenario, it’s not. The software works in those environments, as well as any company procuring supplies and services, including software companies like Box, which Arkestro counts as a customer.

The company, which was originally called BidOps, launched in 2017 after Zagorin saw a need for this type of software while working as a procurement consultant. The company currently has around 50 employees, and is aggressively hiring.

As he expands the startup, he says building diversity is an important factor in hiring. “We do focus on building a diverse and inclusive team. I think that being geographically diverse and remote and flexible helps with that. We have a major engineering hub in Atlanta, which we plan to continue adding to, and we make it a priority to consider diverse candidates for every role that we hire,” he said.

Today’s round was led by NEA, Construct, Koch Disruptive Technologies (KDT) and Four More Capital, with participation from seed investors Cervin, Correlation, El Cap, Forum, Illuminate and Tenacity.