Ever thought of a keyboard that also thinks for you? A small app that sits in your phone and captures your intentions and helps you discover things (solutions/offers) around you. We are talking about a revolutionary keyboard application – Xploree – it predicts fun recommendations and discoveries for you even as you chat with your family and friends. Xploree already has more than 5 lakh downloads among smartphone users. Here are the edited excerpts of an interaction with Sunil Motaparti, CEO & Director, KeyPoint Technologies, where he explains how a tragedy led to the birth of the company and the development of the Xploree app and how it helps users.
Q1. What is predictive search technology?
Predictive technology as the name indicates is about predicting what the user intends to look for, to subsequently offer recommendations that are relevant and satisfactory to the user’s needs. We do so by analyzing user conversations as and when they type an input which indicates additional context for which they would most probably like to get discoveries as results (product offers/content/local content etc). So, before the user actually starts looking for some relevant content, it is already made available to him/her.
Q2. How is the predictive search different from user initiated search?
As mentioned in my earlier answer, the advantage is that content is already fetched and there is no need for the user to go and specifically search for content, in an app or via a search engine. This also saves the user from having to plan and perform a specific search query to get what they want.
Q3. How did the idea of Xploree take birth?
The genesis of KeyPoint Technologies lies in a rather tragic incident faced by one of the founder promoters of the company. Our founder Sanjay Patel’s brother lost an arm in a motorcycle accident. To help his brother use the keyboard easily again, Sanjay started working on ways to improve upon the keyboard ergonomics and input experience. This led to the establishment of KeyPoint Technologies. In addition to ergonomics, KPT has continually worked on improving input linguistics and software from the very beginning. Even though Sanjay himself has since exited the company, KPT carries on with the legacy of combining linguistics and computing technology to improve the text input experience across all types of connected devices. Xploree is our next level of offering to consumers which combines predictive technology with serendipitous discoveries for delightful experiences.
Q4. How does Xploree work?
Powered by Natural Language Processing (NLP) and Machine Learning, Xploree comes in the form of smart multi-lingual keypad. The Xploree keypad understands, interprets and predicts smartphone users’ intent with their tap on the keyboard. It then connects users to relevant brands, offers and content even before the users can realize and verbalize their own intent. The Xploree options appear right on keyboard as non-intrusive ‘Intenticons’ even as the users converse or chat across apps. Xploree brings relevant discovery experiences to users in their exact moment of need, thereby setting new benchmarks in the consumer engagement and experience.
Q5. How does it benefit the users?Xploree keyboard works across apps. So, users do not need to leave the current page / site / app to access another in order to fulfil their needs.
Xploree keyboard works across apps. So, users do not need to leave the current page / site / app to access another in order to fulfil their needs. Xploree works on intent prediction, matching it up with relevant offers. So only relevant offers are made known to each and every user, each and every time; that too discreetly without causing disruption to users’ chat / conversations. Users do not get hassled by irrelevant offers. Xploree offers predictive solutions to its user even before they get to realizing, verbalizing and searching solutions for their intents and needs.
Q6. How would you predict the future for predictive search technology?
With the evolution of Chatbots, NLP and AI, lot of applications have started to embed this technology within their environment. Also, there is a huge push towards defining a common schema (Refer http://schema.org) of different entities (products/services/offers etc) to enable predictive search work better. Particularly on the mobile domain, search engines built on old classical models, will definitely take a back seat. Predictive search is going to save lot of time for users by searching for content based on the understanding of the user’s behavior, needs, wants etc, and making it available instantly. Google Now/Google Assistant, Grokr (for iPhone) are very good examples of what predictive search can do.
Q7. What is hyper contextual advertising?
Hyper-Contextual advertising is about targeting advertisements by identifying additional context (not just keywords) using the mobile device and associated sensors (like location). For example, the choice of serving a Café Coffee Day ad versus a Starbucks ad could be dependent on the user’s current location, their specific tastes, consumption behavior in the past, additional information in the conversation like availability of free WiFi etc.
Q8. How do HCA and PS benefit the marketers? How does it fare with regular marketing and advertising processes?
This helps marketers in more specific targeting because of the additional context. Also, showing up on the user’s device as soon as the relevant intent is expressed, without having the user mention it explicitly is definitely of very high value
Q9. What are the key drivers to the success of this technology?
The success of this technology depends on two things
1. The accuracy in our predictions of user needs and wants
2. The relevancy on the discoveries that we serve to the users
Q10. What are the challenges you are facing in this and what are you doing to tackle them?
A. The challenge is with respect to the ability to quickly iterate over several ideas and on-boarding new partners. We are expanding by adding more resources both on the technology side and the sales side to make this happen quickly. Other than that like with any new technology/product, there is always the challenge of user adoption, increasing the user base, facing new/existing competition, etc.
Q11. How do you predict the future for this technology?
This technology is already being adopted in many applications in the name of chatbots. Some of the examples would be messaging apps, customer service apps, content discovery apps etc. Not to forget the System level offerings like Google Now, Siri, Amazon Echo etc. Also, with the NLP and AI technology being made available as SDKs and libraries, lot more apps will get into predictive search and discovery space. Over the next five years, contextual discoveries, intent-driven smart utilities and branded bots will become the next wave in online marketing. The meaning of localized discoveries will also evolve from a proximity driven recommendation approach and get closer to the consumer mindset.