Imagine you’re preparing for a business trip. Instead of searching furiously for the email with the exact departure time buried in your inbox, and then looking up the weather at your destination, you simply say “Alexa, what time does my flight leave? And what will the weather be like?”
Assuming you have an Amazon Echo, Alexa will then call up all the requested information about your trip so you can arrange for your taxi at the right time and pack appropriately.
According to Gartner, conversational systems are one of the top 10 technology trends for 2017. While the technology is still maturing, such a solution isn’t too far fetched. A small team of technologists in Bellevue, Washington has developed just such a solution – albeit as a prototype.
Concur, acquired by SAP in 2014, creates software that automates all aspects of business travel from booking to expenses. Because it is cloud-based software, they can open APIs – or doorways – that integrate with conversational technologies like Amazon Echo or Slack.
Concur Labs is doing just that as part of their efforts to bring the travel application to where people work and are most comfortable.
“A lot of people use our services once or twice a year and it’s like re-learning the application each time,” says John Dietz, head of Concur Labs. “So we have been pursuing conversational interfaces – both voice and text – as a way to bring services to users.”
Earlier this year the team released a beta Concur Bot that works in Slack, the collaboration tool that lets teams talk to each other via a text message-like interface. The bot embedded in Slack connects directly to Concur so you can ask about your next trip through the app – whether it’s on your desktop or mobile device.
John explains, “Someone at the airport can type an airplane emoji into their mobile Slack app and it will tell you your gate number and when you need to leave. This started as a beta with a few customers and has now graduated to become a core Concur offering. Some of our larger customers are very excited about it.”
Bots can also help people submit and add expenses more easily or call up an inventory of their expense reports.
Concur Labs also brought its travel app right into Microsoft Office 365 via bots. John says, “This will let people process a receipt sent to their email account from mobile payment vendor Square — as well as Amazon, eBay, etc. — so it goes straight into Concur and automatically populates tabs like amount, vendor, expense type, date, and report. We’re pretty accurate.”
Where Natural Language Processing Meetings Machine Learning
While Concur Bots are already doing pretty well, they are always getting better – in part due to machine learning. Conversational technologies are a subset of the larger artificial intelligence and machine learning technologies. Concur has developed their own machine learning algorithms to help make these conversational interfaces smarter all the time. Says John, “We can track the mistakes and learn from those mistakes – these are all guided by machine learning technology.”
He adds, “machine learning adds probability based on its vast experience of reports and trips it’s seen. An expense report should be black and white, but never is. Like, ’Should I expense the drinks I had while waiting for my delayed flight?’ Machine learning can help make those decisions.”
These are some of the interesting things happening at Concur Labs. And point towards what road warriors can expect in the not too-distant future.