KeyPoint COO tells how AI brings ‘discovery’ to Xploree keyboard

According to Prima Dona, Keypoint is the only player globally to build the AI, NLP flavor that works with a keyboard

Prima Dona Xploree

Author

A tech savvy humanBOT, Sharmistha is a professional writer

At a time when the world is becoming app-centric, the variety of choices in mobile apps never seems to be a problem. But it seldom occurs to us that the ‘keyboard’- an integral part of our app experience- can offer more than just an input method.

Xploree, a user-initiated ‘search and discovery platform‘ developed by Keypoint Technologies predicts the needs of users and instantly connects them to relevant and hyper-contextual deals, offers, products and services. Prima Dona, COO, KeyPoint Technologies in an exclusive interview with Techseen explains how the smart keyboard uses predictive learning to make multitasking on mobile a cakewalk.

Techseen: Xploree has partnered with brands like Zomato, Flipkart, Amazon, Firstcry, Coupondunia and Ixigo. What benefits do you provide to these partners? Since it’s a keyboard app, what difference will it make for the customers using the above mentioned apps?

Dona: Xploree has partnered with these companies to help consumers with the fastest hop to the relevant product, services, content from these apps. When an intent is expressed by a user to buy a refrigerator, or dine out, we would like to handhold the consumer to the right point from within our partner apps, to improve the conversion ratio (making a transaction) or visits or engagement. Our technology simplifies the complexities of AI, NLP and Language and delivers simple discoveries for the consumers.

Google recently launched improvements to their keyboard highlighting the need for the keyboard to help users do a lot more than just typing. We already envisioned the need 2-3 years back, and have steadily built a lot more intelligence into giving consumers the ability to consume relevant discoveries based on the intent expressed.

The customers of our partner apps can firstly benefit from shortened interactions to get to the product of their interest when an intent is established. A lot of these individual apps are trying to solve the problem of user experience when they open the app. A lot of these are based on past usage, purchases, lookups, 3rd party data etc. But with Xploree, the partner apps become a bit more contextual for the customer, where the entry point within the app can immediately reflect what the customer was intending to buy, book, look up etc. A lot more predictive!

Techseen: How can we differentiate Xploree from other keyboard apps like Swiftkey and TouchPal?

Dona: Keyboards with language prediction have been around for a while, but we now have a welcome change in sight. It’s been long since any new innovation has made its way to the all-pervasive keyboard, and that’s where Xploree takes the center stage. Unlike Swiftkey and TouchPal, Xploree adds up as a full-fledged tool which makes multi-tasking on the mobile a breeze.

We have taken the learnings from the keyboard and language technology space, and built a whole new technology of AI, NLP, Discovery. Xploree is an intelligent platform that can analyze text, identify intent and related information, map that to relevant discoveries. We have applied this technology back into the keyboard.

As such, keyboard and prediction technologies were pioneered by us long before the enlisted players. Our prediction technology was pre-embedded on the HP handheld and Symbian devices in their glory days. With Android, the keyboard market opened up a bit more for few players and interesting innovations have come up in this space.

We are the only player globally to build the AI, NLP flavor that works with a keyboard to intelligently predict a lot more than standard text prediction. The Xploree keyboard brings users face to face with cab, food, entertainment, utilities, solutions and many similar bagful of discoveries while they continue chatting (the “mobile interlude” as we call it). It is the best choice for those who wish to do away with constant sifting of apps whenever they incur a need.

Techseen: SwiftKey provides statistics and heatmaps that analyzes a user’s typing behavior. Do you provide any similar reports or stats?

Dona: Xploree has a lot of underlying intelligence to improve user error patterns when typing. We do no explicitly provide heatmaps and stats to the user currently. We have the ability and can ship that offering basis the needs from our primary customers, the OEMs.

Techseen: Can you describe the challenges you face or have already overcome in establishing a predictive platform that has to cater to individual needs?

Dona: We have already come to a point where the phone now occupies a significant chunk of our time even outside our real needs. As users take to their smartphone to stay updated, get entertained and chat endlessly, the possibilities to serve them in the most optimal manner are largely unexplored.

There lies an opportunity as well as a challenge in deciphering human intent expressed on the smartphone in these moments where users exhibit no specific agenda. In such a scenario, it is therefore necessary to look at the utility of smart discovery, the technology changer going to replace much of what we know as search today.

Techseen: In 2009, you launched an interesting feature called “Scan Facebook” for Adaptxt that learns words from the user’s Facebook profile. How does it work? Does this make a user prone to marketers by letting them access his activity on FB?

Dona: We had this in our product Adaptxt to help the text prediction technology learn the patterns of word associations, context and usage of the user on social apps. This is on device learning with no compromise on data privacy. This has nothing to do with marketers or their access to our users.

Techseen: Last September, KeyPoint introduced Intenticons to the Xploree keyboard. How is this feature beneficial for users and how has been the response so far?

Dona: Intenticons are the primary interface telling users about the availability of a suitable discovery that matches their intents. This keyboard intelligently identifies a cognitive response to the users’ needs, meaning it shows up icons that denote deeper thoughts. This can mirror a need to search for a florist, order for a pizza, get a cab, among several others.

For all the intents churned on the Xploree keyboard, response from customers to checkout some of these icons have been stupendous and stand in their millions.

Techseen: What is one specific way in which predictive analytics actively is driving decisions?

Dona: Let’s look at the case of how e-tailers tend to suggest additional products to be bundled as a package deal. They do this knowing how customers tend to behave on impulse.

In the future, content consumption will find a great deal of traction for understanding consumer churn through predictive analytics. The technology through continuous innovations on platforms like the mobile keyboard can make a difference here through preference and intent based predictive algorithms. Xploree’s intent sensing technology is capable of knowing user preference to smartly show relevant content when the need arises.

Techseen: What are your future plans in India? Do you intend to build any alliances with e-tailers, or merchants in the near future?

Dona: Our alliance with the country’s top e-tailers are paying fruitful dividends by helping users contextually. There is a lot on cards in terms of the scope offered by Xploree’s intelligent keyboard which can benefit e-tailers and merchants immensely.

Consumer intent to shop among the several categories of products expressed in the many myriad ways still goes largely unnoticed. What we offer to our partners is the chance to become a part of the larger consumer journey.

Techseen: What are your expansion plans in the APAC region?

Dona: Given our expertise in linguistics and machine learning, we are indeed staring at an attractive opportunity ahead of us. As a part of our well chalked out plan, we will first look to consolidate on building our user base in India and then expand to other regions in APAC.

Cloud based gaming platform Roblox raises $92M
Will Paytm's recurring billing service help SaaS-based startups?