//

If you’re a smartphone user, AI has already entered your life. Most probably through chatbots or virtual assistants – Apple’s Siri, Amazon’s Alexa, Google’s Assistant, Microsoft’s Cortana, Facebook’s M, or any of the countless Magic clones and other AI helpers. The tech titans are on a landgrab to build AI ecosystems – going beyond the boundaries of devices like iPhone, Echo, or Pixel.

But in India, the virtual assistant war has been playing a slightly different tune, using a combination of AI and human help on text-based chat. Chatbots here are trying to iterate and find survival strategies to avoid the fate of many of their counterparts in Southeast Asia who choked to death due to meek adoption, failed transactions, and a money crunch.

Some are integrating with big brands for monetization, some are trying premium assisted services on a subscription model, and some have become bot-builders. Now, one of the oldest of them has made a key component of its chatbot tech open source in a bid to hook developers and users.

Today, India’s first chat-based virtual assistant, Haptik, threw open one of its proprietary tech components – Named Entity Recognition​ (NER) System – that helps its bots decipher conversations better. For example, if you say: Send Pehu one Ruby Woo lipstick from Mac tomorrow, the AI powering the assistant should detect the name of the entities in the block of text and annotate it. So with NER, the earlier command would be annotated quite like this: Send [Pehu]{Person} one [Ruby Woo]{Product} lipstick from [Mac]{Brand} [tomorrow]{Time}.

The response to your request will depend on the strength of the NER; in other words, on how well the system recognizes these key words for what they are.

Haptik, which has been working on it since 2013 – well before chatbots were cool – now tackles over 5 million chatbot requests in a month, its co-founder and CEO Aakrit Vaish tells me, suggesting that its AI has built some serious tech muscle.

Why open source?

“If you have to build NER from scratch, it would take years of data mining. Instead, developers can simply download, install, customize, and use our NER for free to build and power their chatbots,” he says. “Be it for personal assistance, ecommerce, insurance, healthcare, fitness, and so on. Any tech product that runs machine learning will need NER.”

Vaish founded Haptik with Swapan Rajdev in 2013, and launched the app in 2014. Like YC company Magic, AI and real human beings help Haptik get routine jobs done for its users on chat.

Haptik’s NER code is hosted in this Git repository. “Any developer can pick a chatbot-builder like API.ai or Gupshup, start building their own chatbots, and cut short their development time by using our NER plugin to detect all the keywords,” Vaish explains.

“We are the first company anywhere in the world to open source a key part of chatbot tech, specifically this NER piece,” he adds.

So, why open source such a vital cog and lend muscle to potential rivals? “There’s a lot going on with chatbots but it is still very much Day 1. Developer tools and open source technology play a key part in the evolution of any platform,” Vaish says. “Our broader goal is to make chatbots ubiquitous. We believe that everybody in the world, at some point, would be using chatbots to get things done.”

But how will this open sourcing move help Haptik?

“The biggest reason to open source any technology is that you hope everybody
else using it will make it a lot better for you as well. What better way to improve your technology than have all of the world’s developers work on it! Once others start using it, they will have feedback, they will talk about what works and what does not. That will ultimately improve our own technology,” Vaish says.

The starving Magic clones

Toward the end of last year Tech in Asia reported on how Magic clones in Southeast Asia were dying. YesBoss in Indonesia, HeyKuya in the Philippines, and Bemalas in Malaysia to name a few.

All of them operated as a virtual concierge service – you could text a request to its chatbot and it would use its army of human assistants to get it done. The lack of focus on a particular task or a range of tasks was one of the reasons that choked these startups. It also needed to connect with multiple vendors – which were mostly conventional businesses slow or reluctant with tech adoption – to manage the supply side of operations. Coupled with these were the monetization woes.

The pain was felt in India too where chatbots and personal assistants became the new cool in the last couple of years. Haptik was the first, but many others followed – Niki, Wishup, JoeHukum, GoHero, MagicX, and so on.

To begin with, these virtual assistants focused on a set of tasks – such as hailing a cab, booking movie tickets or flight tickets, topping up phone credit, paying utility bills, and so on. While this strategy helped in the success rate of tasks, it didn’t do enough to help these startups make money. The reasons are intertwined.

There are niche startups razor-focused on helping users with each of these tasks. BookMyShow, for example, is a popular startup that lets users buy tickets online for movies, concerts, and other events through its website and apps. Founded in 1999, it is one of the few consumer internet businesses in India that is profitable. When it comes to ride-hailing, there is Ola and Uber; Paytm and FreeCharge for utility payments, mobile top-ups, and so on. Most of them had established their value propositions and the tech-savvy app users had got habituated to them before chatbots arrived in India.

“Will chatbots be faster than let’s say Uber or BookMyShow apps? Probably not. But what you really want to compare is the ease of use. Our belief is that things like a screen overloaded with lot of buttons, maps, and other such elements make it intimidating and complicated for some users. The chatbot has a much easier interface on the eye. A few messages and things get done,” Vaish says.

His thesis runs like this:

• There will be a set of users, hopefully more and more over a period of time, that will prefer the chatbot interface for its ease of use.
• Secondly, a chatbot or virtual assistant like Haptik will be useful for infrequent services. Vaish says most people will have two or three things they will do extremely frequently on their smart devices. It makes sense to have apps for each of those.

“For example, I take flights and cabs very often but I watch movies or order food less frequently – once or twice a month maybe. I pay my phone bills once a month only. So if you think about all the other such infrequent needs, the combination of them – across its 500,000 userbase – become a frequent use case and that’s where it is better to have a service like Haptik,” he says.

See: Humans are optional: bots can now gang up on their own to become superbots

Where’s the money?

Haptik has raised about US$12.2 million in funding from Times Internet and Kalaari Capital. When it raised its last round of funding – series B led by Times Internet last year – Haptik had said it was processing half a million requests each month.

Times Internet is the digital arm of Indian media conglomerate The Times of India Group. It claims to reach over 150 million unique visitors a month. The investment from it seems to have given Haptik’s user base the expected boost.

Currently, it has half a million active users and processes over 5 million chat requests monthly, according to Vaish. Haptik works with 22 vendors – “partners,” Vaish corrects me immediately.

“Whenever you make a transaction on Haptik, we take a small fee from the merchant,” Vaish explains.

Another one of Haptik’s monetization strategies is to work with big brands in the finance and consumer goods sectors for deep integrations.

In March this year, insurance company HDFC Life announced its chatbot powered by Haptik. “Such branded integration is going to be one of our three ways [for monetization],” Vaish says. Over time, this could be Haptik’s biggest revenue stream. HDFC Life’s bot will chat with users, help them pick suitable insurance plans, and answer queries. It runs entirely on Haptik tech.

Haptik’s partners and customers include Goibibo, Uber, HDFC Life, Coca-Cola, Samsung, and Amazon.

The third monetization channel will be premium assistant services for consumers. Vaish adds: “Where the users would sign up for some sort of subscription service where they will be able to pay and get some additional benefits.”

Concierge bot Wishup, backed by 500 Startups, pivoted to the paid model early this year. “Unless you are driving a good volume of orders to the merchants, you can’t really charge a commission,” Wishup co-founder Neelesh Rangwani tells me. Today, Wishup calls itself a “lifestyle manager.” You have to fill up a form to tell Wishup of your requirements, get an invite from Wishup, then pay for a subscription plan, and get a human assistant who will work for you remotely.

“This model is working out and we are expanding,” Rangwani says.

JoeHukum, on the other hand, has pivoted to become a bot builder for businesses, hosting and managing bots as well as offering analytics on top of it.

Meanwhile, Niki is beta-testing a voice feature that will let users ask Niki to shop for products and services. “To help shape it better, we are looking for beta testers who can talk to Niki and point out bugs, or suggest improvements,” the startups says in its latest blog post. The war of voice assistants is already being fought by the tech titans: Apple’s Siri, Amazon’s Alexa, Google’s Assistant, and Microsoft’s Cortana. It’s worth watching what Niki, with its team of techies from some of India’s top tech universities, brings to the battle.

Both Niki and Haptik won the AWS Mobility Awards 2017 in the deep tech category.

This post India’s chatbots pivot for survival in the age of Siri and Alexa appeared first on Tech in Asia.

from Startups – Tech in Asia http://ift.tt/2sH7BZT