Sourcing Bots Bid to Keep Freight Moving
One of the unanticipated impacts of COVID-19 is a significant increase in demand for shipping. To help organizations source on-time deliveries of goods amid the chaos, a company called Keelvar has created AI-powered bots that can replicate the activity of human shipping buyers, including bidding for coveted spots in limited freight lanes.
For the past 10 months, shipping companies have been swamped by a surge in demand for delivery services, thanks in large part to the COVID-19 economic lockdown and the reduction of in-person shopping. In July, UPS said it was “delivering volume similar to peak holiday shipping season,” while an October survey by JOC (the Journal of Commerce) found that nearly 88% of LTL (less-than-truckload) shippers says longer transit times is an issue.
This surge in shipping demand has made it harder for companies to ship goods, especially when the available capacity is down. According to Accenture, air capacity is 17% lower this week than it was a year ago.
The increased demand and lower supply results in a substantial increase in bidding activity for a limited number of freight lanes. That means that employees must spend extra time to ensure their companies meet their own logistical goals.
This is where Keelvar steps into the picture. The Cork, Ireland-based company has created “intelligent sourcing bots” that can execute of the tasks that employee regularly handle when it comes to preparing air and ocean-going deliveries, but without requiring human workers to put in all the time themselves.
According to Alan Holland, the founder and CEO of Keelvar, the sourcing bots work similarly to the familiar chatbots that greet visitors to popular websites, with one key difference: They’re working in the highly specialized field of air and ocean freight.
“Whereas simple chat bots on e-commerce sites often just search for support articles, a sourcing bot is more powerful in that it can take an instruction to do a multi-phase job, figure out how best to do it, and revert for further guidance at key control points,” Holland tells Datanami.
Keelvar’s AI-powered freight bots mimic the work that real human agents would do to prepare a given shipment for delivery. They are Level 4 bots, which means they don’t work completely autonomously. But when overseen by an individual, they can significantly reduce the amount of manual work that would ordinarily be required of the human.
“Bots are able to establish spot bidding events within minutes, managing and automating tedious tasks such as inviting carriers, collecting and validating bid data from carriers, messaging bid status updates, conducting necessary rate card and lane information lookups, and generating award recommendations,” he says. “As a result, companies can increase competition and savings, improve decision quality, increase process compliance, traceability, and ensure operational continuity with less risk.”
Last week, the company rolled out its air freight sourcing bot, which joins the ocean freight bot that was already available. The company says it early adopters have already run nearly 100 air freight sourcing events through the bots over the past several months.
Keelvar specializes in developing software solutions that help organizations optimize their sourcing, procurement, and fulfillment activities. The company claims that, through AI and machine learning, it’s able to optimize the supplier choices that customers make. It claims to have saved clients over $100 billion.
Holland founded Keelvar in 2012 after leaving University College Cork, where he was an AI lecturer in the Computer Science Department. Holland’s specialty was optimization, game theory, and algorithmic mechanism design.
Holland sees AI bots becoming more prevalent in the sourcing world. “Throughout my career, I’ve witnessed organizations with hundreds of personnel running routine auctions/eRFPs and spot bidding, and invariably executing them sub-optimally in choices for bidders or with limited feedback to drive competitive tension,” he says. “Sourcing excellence requires precision and attention to detail that is difficult to scale via human-operated processes. The human + bot model is proven to produce the best results.”