There are few things more convenient than ordering a Uber via Twitter or requesting one’s bank statement using SMS, and actually getting it.
One of those few things is doing both these using day-to-day language. Something along the lines of typing, “kindly get me a Uber ride from Menlo park to Woz way” on Twitter to a Twitter handle and getting an actual cab. We have not been able to until do this…until now that is.
A Kenya-based enterprise solutions company, Ent-Mobile, is attempting to do this via Pythias, its flagship product that allows users make service requests to companies using natural language.
Ent-mobile CEO, Andrew Kamau, told TechCabal that the idea is to simplify “the digital service delivery model.”
In a pitch at the recently concluded Citi Mobile Challenge finals, Andrew said the product will make service requests as intuitive as chatting with a friend by using platforms that customers are already familiar with, compared to models that require pre-configured portals and keywords.
“We believe that the current model is intrinsically broken. Users have to frame their service requests to conform to what computers can understand, so to speak,” Andrew told TechCabal, “We believe that computers should understand us; where we are; how we speak. This is our aim. We do this by bringing the services to platforms where users are (social media and mobile platforms) and in a channel akin to talking to a friend – natural language).
Pythias uses a natural language processing algorithm that converts natural language into a service request on the business side from where pre-configured responses can be effected. The product also relies on multiple APIs to facilitate service requests. The product uses IBM’s machine learning and cognitive infrastructure, IBM Bluemix and Watson to effect the conversion.
In the pitch where Andrew demoed a cab-ordering use case, the product leveraged on Google’s geocoding API to retrieve location and Uber API to send the cab. Pythias won the award for the “Most Innovative Use of Bluemix” at the competition.
There are enormous use cases, but Ent-mobile is targeting the retail banking with the current iteration.
Pythias is already building traction. “ We have deployed it with a virtual office space retailer in Nairobi, VOS,” Andrews tells me. He says Pythias is also integrated into other clients who can’t be named now, but will be made public soon.
The startup was incubated at the Nairobi-based iBizAfrica.
Photo Credit: Warm ‘n Fuzzy via Compfight cc
