If some clever Eastern Europeans have their way, parts searches will become a lot easier, faster and more routine. Procurement staff or others in need of a part can just tap out their needs in an ordinary e-mail on a PC, laptop or mobile device, and an efficient tool will provide several options in a few seconds.

Locatory.com has just launched an artificial intelligence tool, Amber, that translates e-mail inquiries about parts into searches for the right item, condition and quantity; replies to the inquirer with availability and suppliers; and then can automatically send a request for quotation (RFQ) to each supplier, all in less than 15 sec.

Where did it come from? “We are constantly generating ideas on how to ease the procurement and logistics processes in the aviation industry,” explains Locatory.com CEO Zilvinas Sadauskas. “Eventually we came to a conclusion that the most common interface, used by most of the aviation supply chain members, is e-mail. Everybody understands how it works and how to send one.”

Still, Sadauskas says he is unaware of anything similar to Amber in aviation trading. Feedback from beta testers has been positive and surprised, some asking: “How is this possible?”

Amber was built entirely in-house by Lithuania-based Locatory.com’s own IT team. The software assumes all incoming e-mails are requests for parts and looks for desired part numbers, conditions and quantities.

The e-mail must thus contain at least a part number. “You can write whatever you want, and Amber will pick out part numbers from context,” Sadauskas explains. “You can even forward an e-mail received from your planning department or from a customer having an AOG [aircraft on ground], and Amber will do everything for you.” Or a buyer can simply copy Amber when sending an RFQ to his regular RFQ mailing list.

Sadauskas says Amber is a lot easier and more natural to use than asking procurement stafffers to fill out standard digital forms. “People are looking for the simplest solution they can get and Amber is as natural as communication with your friend.” For example, one could hit the send button on the following message:

 

“Hi Amber,

I’ve just received an e-mail from an airline I’m working with. Their window has just cracked and they have an AOG at the moment. Therefore I would like you to tell me who has the part number MY20266-10 in overhauled condition.

Please deliver me a report as fast as you can.

And don’t tell me that you’re tired— you’re an artificial intelligence, right?

Cheers!”

 

Sadauskas says the tool works for any part. “Amber technology is based on artificial intelligence engine. That’s what makes it so elegant and fast.”

E-mails do not need to be written in English. The tool has been tested in several languages, including Lithuanian and Russian. Amber even understands Cyrillic terms for Russian-built aircraft.

Given the variety and freedom of communications allowed, Amber might misinterpret a few messages. It looks for structures that resemble part numbers. “Imagine a Chinese child looking at Latin characters or a Lithuanian child looking at Chinese characters,” Sadauskas explains. “These characters look like images without meaning.” Some images look like part numbers, some do not.

Nevertheless, Amber gets its right 99% of the time. In the other 1% of cases, users can log in to the standard Internet or mobile interface of Locatory.com and search by keypad or voice query.

And even these few errors should diminish over time. Locatory.com staff will teach Amber and help her understand. “Amber will not teach herself,” Sadauskas acknowledges. “But our IT department will constantly iterate on the errors.”

When the potential buyer receives Amber’s summary of availability by part, region, quantity and condition, he or she can just hit “reply.” This sends a standard Locatory.com RFQ to the suppliers.

Sadauskas says Locatory.com’s quality-assurance policy prevents competitors from using Amber to investigate each other. Amber is available to customers, adding: “If one wishes to become our customer, he has to comply with our quality assurance policy.” Part documentation is handled by separate communication between buyer and supplier, independent of Amber.

Sadauskas is eagerly awaiting market reaction to Amber. He hopes it will help optimize the procurement process. “We have plans to build a Web service for Amber, for buyers to be able to integrate Amber with their internal processes. Can you imagine how many copy-and-paste procedures can be saved with that?”

Amber probably can. 

Editor's Note: A previous version of this article described Amber Technology as being neural network-based. It is based upon an artificial-intelligence engine.