Interview

We, at NextBigWhat, are attempting to drive forward the conversation around Artificial Intelligence in India beyond headlines that scream for attention but don’t offer much, buzzwords that make your head buzz after a while, and countless utopian tales that vie for your eyes only to leave you with a scratching head and a lingering itch to actually get to the meat of the matter.

In this spirit, the #AIBoss series intends to educate, inform and elevate our readers understanding of the subject as well as the space. If you’re part of the leadership team of an AI based product, from enterprise or startup, willing to share your AI learnings – then reach out to us and become part of this initiative (you can apply using this form).

Adarsh Natarajan is the CEO & Founder of Aindra, a medtech company that offers AI based computational pathology solutions. Their first product, CervAstra, is aimed at automating the detection of cervical cancer, an illness which causes more deaths in India than in any other nation. Aindra says that their core computational pathology platform, Astra, has been designed to be extended to various other critical illnesses including, but not limited to, other types of cancers.

We caught up with Adarsh for a comprehensive tête-à-tête about Aindra, the kind of work they’re doing, how AI can help solve healthcare woes in India, the challenges that exist in AI diagnostics, and much more.

How would you introduce Aindra?

So, to put it a little technically, we are an AI-based computational pathology company which we leverage to provide point-of-care diagnosis for critical illnesses. In layman terms, we provide the ability to detect certain kinds of cancers right at ordinary physician’s clinics.

So, from scanning the history of Aindra, it seems that you pivoted from education-centric AI solutions to AI Diagnostics. So how did you arrive at the decision and what prompted the change?

Initially in the AI space we were looking at completely different use cases. We were going to do identity management in the education space, and then we pivoted away from the education space and started doing stuff for other sectors with broader positioning. We started looking at computer vision much more broadly – not just face recognition, which we started with – and I had a moment where I realized that it has tremendous application areas.

Technology like AI and Computer Vision have vast, multiple application areas. That’s when I started meeting people from across sectors and domains to talk to them and understand what is happening in their spaces. I met people from the automobile industry, manufacturing, textiles, and then also in healthcare – and healthcare struck a chord with me.

There were much more lucrative and easier routes in other segments, but what struck me was that we could democratize technologies like this and make it available for a large population to solve big problems in an economically viable manner. I felt passionate about that.

And so, we zeroed in on cervical cancer as the first condition to take on, while realizing simultaneously, that this would probably be our first beachhead and that the core technology would be leveraged to tackle other conditions going forward.

What were the initial challenges you encountered when you began building a solution for the clinical pathology space?

I come from a technology background, and we started with building only the AI software part as proof-of-concept and tried it out on small data sets. But there were absolutely no enabling systems available which would help the AI…

You mean integration to facilitate the transfer of data from slides to the system?

Yes, but there weren’t even any systems to capture the data first. You need to capture it, then convert it into a digital format and then apply the AI algorithms. So we built the AI algorithms and were quite gung-ho about it and then when we started talking to customers we realized that we hadn’t done a good job in assessing the limitations.

We wanted to provide the ability to detect critical cancer conditions not only in tertiary care oncology centers in cities like Bangalore, Mumbai, Chennai and Delhi but also make it available to primary physicians or clinics in faraway Tumkur, Bijapur, Ballari – tier three towns.

If that were to happen, we had to build the enabling systems too and backwards integrate them to create an end-to-end ecosystem. That meant building hardware which just wasn’t available commonly. The bitter truth hit home that India still imports 75-80 percent of its medical devices from outside and most aren’t built for point-of-care but for centralized large centers such as a Vedanta or a Fortis. So, we had to design everything from the ground up to take the bull by the horns in this space and truly create a point-of-care system.

To build the ecosystem of the product, so to speak…

Yes. What started as one bespoke application morphed into a solution with multiple components all working in tandem to act as a single end-to-end product. Which meant we had to deal with designing hardware and getting them into manufacturing them and sending them into queue, so it is a whole different ball game than what we had expected.

Tell me a little about the deep learning systems you’ve implemented.

When we started off, deep learning was quite nascent in terms of its maturity, so we began with traditional machine learning techniques. We then started experimenting with deep learning because it has given us a quantum leap over some of the traditional techniques. But even today, it’s not one size fits all. We’ve really had to tinker, experiment and build a suite of algorithms to make it work. Today, we have a mix of deep learning and traditional machine learning to do the processing and then newer techniques like Deep Neural Networks, which brings its own strengths.

We also have to deal with the hard constraint of being in the medical world, where data is a real luxury. It’s not like we’re dealing with consumer data where you can crawl the web and scrape endless amounts of information to feed into your system.

So we’ve used some standard frameworks like TensorFlow and Caffe. We’ve also had to use transfer learning because of the insufficiency of data as I mentioned earlier, and build customized networks to deal with it.

On the clinical side, how did you go about partnerships and sourcing the data?

We had to partner with a very reputed tertiary oncology center called Kidwai, which is one of India’s best, go through an extremely long cycle of approvals before we got the final go-ahead. We then worked with their pathologists and used data from their archives. So we basically co-opted them into co-creating the product and the algorithm. They imparted their clinical expertise and we translated it into an effective AI. It was a long journey beginning with extracting medical data, drawing useful information from it and so on. Digitization of the data was a big and crucial part i.e turning archived glass slides into usable data format for training our system. We are now in the stage of clinically validating many of our solutions which will be done by mid-2019, hopefully.

And I’m guessing all of this data ultimately feeds to your computational pathology platform named Astra. What can you tell us about it?

With the computational pathology platform, what it allows us to do is extend our application area to tackle other conditions by retraining newer models with newer data by leveraging the same layer flow framework. We’ve been looking at other cancers such as blood cancers, leukemias and hope to move from cytology to even histopathology – which is the study of tissues – which means we would be going after prostate cancers and the like.

I was just about to ask you if Astra could potentially tackle detection of organ malignancies and other forms of cancers…

Yeah. I hope that answers your question somewhat. We’ve taken a lot of pain and effort to get to where we are today. And in hindsight, it was worth it because it has now prepared us to replicate it in an easier manner since our learning models can be fed with newer data to provide effective results. As an experiment, we’ve been able to prove this to ourselves by using a completely different dataset that deals with breast cancer tissue and we achieved equally good results as our core solution that deals with cervical cancer. Which means our network is now generalizing well enough.

What are the current challenges in cervical cancer detection? You mention affordability as being one on your site, but in my conversations with a few pathologists I was told that it is per se not a major problem.

There are multiple levels of efficiency and it’s not just one aspect that counts. When we talk about the cost of a test, it’s not just the price of a pap smear test that a woman has to deal with.

Imagine you had to get some kind of a test done but it is only available in Mumbai, it costs a 100 Rupees but you’re in Bangalore. The cost now increases exponentially because you have to spend money on travel, take a few days off from work and so on. For women in India, it could mean having to leave children and family. And if you’re a daily wage earner, then it would mean having to lose your paycheck for a few days. So it all adds up. If the test was available to you closer to where you live or work it would make a major difference. Accessibility to proper mechanisms drives affordability.

Let me give you an anecdotal reference, I have seen with my own eyes a pap smear test report (along with the bill) costing 2000 Rupees at a panchayat away from Tumkur. The same test in Bangalore would cost you 500-1000 Rupees. If you’re a below poverty line patient, and you go to a center like Kidwai, it would probably cost you 200 Rupees.

So you can see how the problem gets acute in an inverted way. Ultimately, the pathologist who actually who is doing the reporting is probably going to be paid the same amount. The center which picks up the sample had to handle logistics to transport it to Bangalore, there is labor involved in a clinic, margins involved and so on. The cost ultimately gets padded up.

So that’s the economics of the problem. What are the other challenges?

Awareness is a major challenge. And that is compounded. How many of us actually go for regular health check ups once a year? Very few. We don’t go to a doctor unless we’re sick for sure or feel a lot of pain. Or if your employer has a corporate healthcare plan so you get preventive health check ups. Or perhaps your insurance provider compels you to get them done. These are some of the triggers. Now imagine you didn’t have these triggers and you had to travel the distance of Bangalore-Bombay for some basic diagnostic tests. That would worsen your willingness.

Factors such as this compound the problem and have led to India becoming the cervical cancer capital of the world. One woman dies every seven minutes due to cervical cancer in India. These are official numbers. And we’ve been talking for longer than that, so you can do the numbers.

Remember, too, that many people who die in India never actually get the right reasons ascribed to their deaths at all.

Your solution then also provides advantages in terms of turnaround time for cervical cancer detection, obviously…

Currently, if you’re in certain tier-II cities or towns that border them, the turnaround time is around 4-6 weeks. The analysis itself doesn’t take that long, but because it has to be picked up and transported to, say, a Kidwai or St. John’s, who are already having to deal with their internal requirement, and then get entered into their backlog.

With a point of care system like ours, the patient comes in, provides the sample and within an hour and a half, during which she could get other tests done, she will have her report ready. This fundamentally disrupts a lot of things.

One of your stated aims is to democratize access to healthcare and this certainly seems to fit the bill…

Yeah. Pap smear testing is not something that we have invented. It’s a proven technique that is well tested and has existed for decades. With proper testing and proper usage, cervical cancer rates can dramatically come down as evidenced by the incidence and mortality rates in the US and UK, where screening or yearly or at least bi-yearly tests are done on a very regular basis.

In India, with the population, lack of accessibility, and heavily negatively skewed physician to patient ratio, you’re never going to be able to fix it unless technology acts as a force multiplier. And I’m not just talking about this particular condition, any condition for that matter.

The US has 23 doctors for every ten thousand people and the UK has about 25. India has 7. So this is a systemic constraint.

And you certainly cannot manufacture clinicians overnight either…

Absolutely. Let’s take cervical cancer, for example, which, in India, has around 350 million women in the risk category. Which is more or less the size of the population of the United States. How can we possibly change the status quo dealing with this kind of scale if we keep doing the same old things expecting new outcomes?

The primary health centers which you’re keen to provide your solution to have their own challenges. Such as lack of electricity, non-availability of round the clock internet and so on. Have you factored those in?

Absolutely. Which is why we don’t have the AI algorithms on the cloud. We haven’t built an exotic or esoteric application with the assumption that there’ll be connectivity around the clock or that there will be electricity round the clock.

This is an edge computing device, where the first level of triaging – whether a sample is normal or abnormal – takes place in real-time. For confirmation, we bring in a pathologist for a report on a non-realtime basis. So once the sample is analyzed by a pathologist, only then is the confirmation provided.

What we’re doing is decoupling the need to have a pathologist available all the time for analysis and reporting. From a medico-legal perspective, we still have a pathologist sign off.

It’s timely you say that. Because one of the questions I had was whether your solutions are built into the hardware, cloud-based or some kind of a hybrid model…

Yeah. Putting it on a edge device is extremely hard because you’re talking about algorithms which require serious computing power. The easy way would have been to put the solution on the cloud, where it would only work in urban centers with good connectivity but that would not serve the purpose at all. So we really had to re-look at every bit in the entire value chain to decide what had to be done.

Are primary health centers the only place where this could be disruptive?

We’re not just talking about primary care centers, we’re also looking at gynecology clinics who don’t do cervical cancer testing because they don’t have the equipment or the resources to do it. They will just refer the patient, if at all, to a pathology lab which itself has a host of problems as I mentioned earlier.

So, imagine a patient walks into a gynecology clinic and the gynecologist has two compact devices on her desktop. The gynecologist collects the sample, does the examination and in a short while the system confirms whether the patient needs further scrutiny or can be sent home. It could be that simplified.

What are the challenges you’ve faced in implementing the AI on your custom hardware?

We had to keep in mind the kind of computing power necessary and figure out the kind of GPUs we need and whether they’re available in the form factor that we desire. Our goal was to put out a model with a smaller footprint device. We had to also optimize how and at what speed the data is processed based on computing ability. And so on. A lot of thought went into it.

Pathologists I spoke with told me that some of the legacy machines throw up red flags in case of ambiguity in data interpretation, which meant that they would have to manually confirm some of the findings. What kind of manual fallbacks have you incorporated?

As I said earlier, at the first level we do a triage result – normal or abnormal – and we indicate areas of abnormality. We present the findings to our panel of pathologists via our telepathology medium and once they sign off, the report is sent back. The findings of our system are presented to the pathologists, as they analyze, to assist them by calling out the areas where it has detected abnormalities, speeding up the process.

I was told that there might be situations where further data is necessitated from the primary clinician…

Clinical information such as the age of the woman, whether she is married or not, or how many kids she has, whether she has had multiple partners, whether she is menstruating or not etc. is already collected without us being in the picture. This is done by the primary care physician and this data, too, is presented to the pathologist if the system finds any abnormalities. And they can choose to override the system if need be.

A number of industries have seen transformation due to the revolution of low cost sensors. Have you benefited from it?

Absolutely. A lot of things we are able to do is because of convergence of a lot of these factors which have driven down the cost of computing dramatically. If that wasn’t the case, we wouldn’t have been able to make use of our technology. And sensors too, which are crucial to our solution. Advances like these have played a large part in democratizing technology. Which is why I believe we’re in a great point in time. If only India had a semiconductor industry, we could’ve achieved enormously. But it’s an imperfect world.

Broadly, what is the response to AI Diagnostics in India?

There’s a lot of apprehension as well as interest right now, depending on who you talk to. You’ll largely find a bit of both. Some people are threatened with the advent of technology and there are a few people who are able to look at trends that are unfolding and much more accepting of the changes. There is certainly a high level of awareness about the oncoming of AI.

What is the biggest challenge in your view in diagnostics/medical AI?

From my experience of having spoken to various researchers, academicians and clinicians, the biggest challenge globally is integrity of data. That is super critical for AI to become a widespread reality. Since we’re all going to be relying on large amounts of data to make our models as effective as possible, we have to ensure high integrity of data. This means ensuring minimal bias in data because biased data in medical AI is turning out to be a massive problem. Sadly many AI companies aren’t even aware of how important this is – it could determine the success or failure of their systems.

(Special thanks to Dr. Neha Ratan B, M.D. Pathology for her inputs.)

Join 12,000+ CXOs who are reading curated AI news via the newsletter

We, at NextBigWhat, are attempting to drive forward the conversation around Artificial Intelligence in India beyond headlines that scream for attention but don’t offer much, buzzwords that make your head buzz after a while, and countless utopian tales that vie for your eyes only to leave you with a scratching head and a lingering itch to actually get to the meat of the matter.

In this spirit, the #AIBoss series intends to educate, inform and elevate our readers understanding of the subject as well as the space. If you’re part of the leadership team of an AI based product, from enterprise or startup, willing to share your AI learnings – then reach out to us and become part of this initiative (you can apply using this form).

Here is our interview with Ranjan Kumar, the founder & CEO of Entropik, a company that specializes in ‘Emotion AI’ – a suite of technologies that allow brands and creatives to truly guage the emotional response of consumers with the incorporation of brainwave mapping, facial coding & eye tracking.

NextBigWhat caught up with him for an extensive discussion on Entropik’s product, the ethics of emotion tracking technology, where AI is headed and how their proprietary technologies are making an impact in various business verticals.

How would you describe Entropik?

At Entropik we work in the space of emotion intelligence. So what we do is basically, track consumer emotion at its point of origin in a scalable way. We’ve built technologies around brainwave mapping, analyzing facial expressions, analyizing eye tracking, so these are the fundamental proprietary technologies which actually allow us to decipher the subconscious aspect of human behaviour. And 95% of the decisions that a consumer actually makes is subconscious, however most of the analytics aligns around the conscious aspect of user behaviour. So we’re taking our shot at tracking the subconscious aspect of human behaviour.

What are the current and upcoming applications for your product suite?

Right now we’re very focused on media experience testing. So anything video – right from ads to trailers to long-format TV shows and even 3-hour films. Where you get to see on a second-by-second level what the emotional response of the user is, be it happy, sad, excited or bored. In addition to that, we also track what the attention level is, where the mental engagement stands, and quickly you can get to know the parts of the video that least engage the viewers.

UX testing is another area where we offer great value. Some other areas are chatbot experience testing and shopper experience testing. So while you’re walking in the store, we can track where you’re looking and what sort of brainwaves are being generated. Whether you’re attentive and feeling excited or bored etc.

Edtech too is emerging as a big market. We’re using the technology to validate psychometrically the cognitive learning index of a student. Retail is at an early stage and automotive is an upcoming market where companies want to track fatigue and stress levels of drivers.

What was the pain point you set out to solve specifically with Entropik in its early stage?

One of the resons why products fail is because they don’t resonate with consumers. Every consumer brand, the biggest pain point or the fundamental question is whether I can decipher what the consumers are feeling, what their preferences are, and hence align and create a product which is likely to resonate at a more cognitive level with the user. The first step to that problem is knowing whether you understand your consumer very well and whether you understand not just the conscious aspect of their expression, but also subconscious behavior and subconscious elements which tie into the buying pattern.

We looked at what makes a purchase decision or what makes a content consumption decision: fundamentally 95 percent of your decisions are subconscious. That means we need to have analytics to pick up on this aspect of consumer preference. So that’s a huge problem.

Look at the kind of losses companies have when they actually roll out a product and it fails. Or if a movie is released and it fails – we’re talking about huge losses. Efficiency of success of product content is around 2 percent today. So we set out to solve this 98% inefficiency problem.

Based on the sweep that you have in terms of what you could do with this kind of technology on scale, could we be heading towards a Huxlian world where everyone’s emotions are fine tuned to an inch of what is being intended to evoke – pleasure, sadness and so on?

I don’t think that will happen. So, you have AI which is about intelligence i.e it is logically intelligent. It has a bunch of ‘if x’ rules, that is ‘if x then do y’. A huge collection of such rules is called artificial intelligence. What it lacks is empathy and emotional intelligence, which means that it is answering a question but without the context of my current emotion.

When two people are interacting, emotion is a factor which makes human communication unique. That aspect is missing with machines. So I think that brings a lot of resonance but I don’t think technology will be modifying people’s behaviour in the way that you envision. I think it is solving a gap of 98% inefficiency of interaction between man and machine. That’s about it.

When it comes to ads, there is an explicitly commercial element to it therefore optimizing for it in the manner that Entropik allows makes sense to a great degree. With movies or TV shows, however, what if production houses begin to edit films based on the responses they see on a system, rather than it being the relatively untramelled creative vision of a director – how do you see that playing out?

That’s something we face day in and day out. We currently work with almost every major broadcaster in the country. And we’ve faced this question before. What our product can do is two things: a. It can help marketeers optimize how they reach customers. Help improve business outcomes. b. It helps creative guys to optimize on the content itself. We’ve been very cautious not to step on the toes of those in charge because the relationship should not trample on their creative vision. We look at ourselves as enablers.

For example, if we’re recommending that these are the 10 things that came out of analyzing this piece of content, then they mostly agree or disagree but still do what they have to do. So that freedom is never taken off from them. And we’re receiving a lot of intelligence from their end as well and we need to make sure that we’re incorporating that into our system. So as to make it even more intelligent. And the other important thing is that most of these studios are sitting on hundreds of scripts, and they require a system that can evaluate these ideas in the quickest way possible – which is where we come in as enablers.

You can watch the full interview in the embed above. Make sure to subscribe to our Youtube channel for more such engaging videos!

We, at NextBigWhat, are attempting to drive forward the conversation around Artificial Intelligence in India beyond headlines that scream for attention but don’t offer much, buzzwords that make your head buzz after a while, and countless utopian tales that vie for your eyes only to leave you with a scratching head and a lingering itch to actually get to the meat of the matter.

In this spirit, the #AIBoss series intends to educate, inform and elevate our readers understanding of the subject as well as the space. If you’re part of the leadership team of an AI based product, from enterprise or startup, willing to share your AI learnings – then reach out to us and become part of this initiative (you can apply using this form).

Here is the second and final part of our interview with Parag Arora, who is a technology entrepreneur with deep insight into messaging solutions, having sold his first startup Plustxt to Paytm in 2013 and currently holding the position of Chief Technology Officer at Glowing.io – a YCombinator selected startup that provides an AI based messaging solution to the hospitality industry.

NextBigWhat caught up with him for a rather extensive tête-à-tête on the journey of Glowing.io from conception to product, AI & NLP in the hospitality space, and whether or not he considers voice based solutions like Alexa a threat in the coming future.

In terms of the tech stack, how does Glowing.io operate? How has it evolved over the product’s lifetime?

The heart of the messaging system is Erlang based RabbitMQ and we have evolved from XMPP based architecture which was also used in Plustxt, later sold to Paytm.

Glowing.io posed new challenges for providing enterprise-grade features as assignment and removal of conversations in real-time manually by staff or automated by our system.

Apart from this, the business layer is built over Ruby on Rails with backend as PostgreSQL. We use native apps for mobile and nodeJS for our messaging infrastructure. For the frontend we use ReactJS.

Two main infrastructures we have developed over time are QA and DevOps automation. We are currently deployed over AWS and Alibaba and both have proved to be very crucial components for us over time.

What kind of machine learning systems has Glowing.io implemented and what do you hope to achieve with it?

Our main focus has been on NLP and further automating tasks inside the organization achieving operational efficiency as an end goal depending on integrations we have done with other systems of that particular hotel.

What are the kinds of analytics you use to measure guest success on the hotel end – as in, what kind of analytics do you make available to them – as well as customer success on your end i.e the KPIs you use internally?

We have a customized analytics dashboard inside the product and the most important component we currently focus on for customer success is guest engagement – which is essentially a percentage of new guest conversations compared to new guest check-ins on a particular date.

With the onset of Alexa and other voice based solutions which are seemingly being adopted in the hospitality industry, do you view them as a threat or as something that you can integrate into your solution? Or even, perhaps, build on top of it?

For us, Alexa is just like any new communication channel. We aim to be the digital hub for hotels for guest conversations.

The only difference we see is that guests are already trained to receive automated instant replies to Alexa compared to messaging channels where guests expect human-like conversations.

Since a product like Glowing.io would presumably entail deep integration into a hotel’s processes – what is the flow like in terms of training hotel staff/personnel to get to grips with the solution? And what kind of timeline do you maintain for it?

We understand introducing a new system with a new process takes time and we give hotels enough time to internally align. Although the overall onboarding can be done in a day, we have divided the onboarding in 45 days with weekly sync-ups to ensure the program is successful from operations, marketing, branding, and sales side.  We also provide an optional onsite training program which I personally love a lot as we get to learn a lot about the property and it provides an opportunity to travel to beautiful places and experience their hospitality.

Developing markets have the advantage of cheap labour – is it a safe assumption that they would rather have a human behind the screen than invest in an AI? And have you come face-to-face with this in terms of sales across geographies? How are you dealing with it?

Interestingly Glowing.io is deployed in equal proportions in developing economies as well as developed. We have never looked at automation as a way only to cut down on cost and resources but rather as a way to be operationally efficient, correct and more data-driven.

According to you, what is the future of conversation between customers and companies in the context of chat, voice and more?

I think the future is that humans will be for emotions and AI will be for automation, taking away the pain and time of redundant tasks. I see a beautiful world of happy employees in hospitality and AI-enabled hospitality for brands who have hospitality in their DNA.

Make sure you’re subscribed to the AI Boss newsletter to keep up with the latest in AI developments across the world!

We, at NextBigWhat, are attempting to drive forward the conversation around Artificial Intelligence in India beyond headlines that scream for attention but don’t offer much, buzzwords that make your head buzz after a while, and countless utopian tales that vie for your eyes only to leave you with a scratching head and a lingering itch to actually get to the meat of the matter.

In this spirit, the #AIBoss series intends to educate, inform and elevate our readers understanding of the subject as well as the space. If you’re part of the leadership team of an AI based product, from enterprise or startup, willing to share your AI learnings – then reach out to us and become part of this initiative (you can apply using this form).

Here is the third and final part of our three part interview (Part I, Part II) with Shailendra Tiwari, founder of Fasal, an Ag-tech (agriculture tech) startup that is focused on building AI and IOT (Internet-of-things) based SAAS solutions to address the woes of Indian horticulture farmers. Fasal’s primary aims are to increase and improve the quality of the yield and lower input costs for the farmer. It is currently deployed in farms across Karnataka, Maharashtra and Chattisgarh.

In this interview, Shailendra gives us an inside view of agricultural tech in India, how Fasal is seeking to disrupt the space and why agri-tech stagnating has more than just a little to do with it involving the sweltering heat in Karimnagar as opposed to the air-conditioned cabins in Koramangala.

Q14. Are you looking at drone integration going forward considering it is being utilized by high-end farms in some regions of the world to disperse water and/or pesticide?

Yes we are. In fact the way Fasal has been architected is to make sure that we can collect data from any source. Be it sensors, satellites or drones.

And let me tell you, not only in Hi-tech farms, I have met farmers with 3-4 acres of land exploring drone solutions.

Don’t forget finding farm labourers is becoming a big and visible problem now. If we can provide drone services at a cost equal to that incurred with that of labours, I don’t see a problem in their larger adoption if these solutions can reach out to the farmers in need.

Q15. It would be safe to assume that apart from key fundamentals, each crop needs its own software and system. What is the research process like and could you share some interesting learnings from that?

Wonderful question. I assure you that each crop/each region has its own requirements. For example there are regions in India where a disease called Powdery Mildew is a big big problem for chilli. But for the same chilli, in a different region, it is practically non-existent.

For us, even deciding which is the next crop we should go after, is complex decision involving agronomy, farmer personas, regions where it is grown, financial consideration, cost of cultivation and at least 10-15 more considerations.

Once we have decided which crop we pick, we do a thorough research on it’s lifecycle, the pests and disease which are a menace to it, the irrigation patterns and practices followed, the amount of research done on that crop across the world, the amount of research done on that crop in India. Post that we identify the biggest problems farmers face with that crop. Once that is done, we do a technical and domain assessment of the feasibility of building solutions around those problems. Once all of this is done, then starts the process of research and development on building and validating the needed solutions.

Q16. Legacy agricultural research institutes have been around in India since the British era and a lot of research work is conducted in India today too – have you been able to take advantage of both the organizations and the corpus of material and/or is there any kind of institutional partnership you’re contemplating in this field? Considering you’re well placed to translate that traditional research into tech solutions.

I absolutely love how this question is framed. These are the same words in the same order I have used at certain places. In fact I was at University of Agriculture sciences, Bengaluru recently and Dr. Chandrashekara, head of entomology department and I discussed this at length.

There is good research in the universities of India. We have very learned agri scientists and professors in India. And as startups we do have the opportunity to take the research done to where it belongs, the hands of farmers.

We at Fasal have been doing it and will continue to do so extensively in future. In fact through this answer I invite any university/scientist/professor who wants to collaborate with Fasal. Not only this, we have also been working on partnering with universities/institutions for the greater good of farmers very very actively.

Q17. Tell us about your tech and AI stack.

Said much already 🙂

Your readers may need to sign an NDA to know more (pun-intended) 🙂

Q18. What can you share with regards to your emphasis on sustainable farming?

To be honest, there is no option but for farming to be sustainable. If we continue with current ways, we will reach a point of no-return. We have to bring awareness that everytime a farmer sprays an extra dose of pesticide on the farm, what ecological damage happens. We have to bring awareness that everytime a farmer irrigates for an hour more than needed, what the world loses and much much more.

I imagine the scenes from Christopher Nolan’s movie Interstellar where Blight has taken over the world and food production is under severe threat and thus humanity’s survival. It may sound far-fetched but if you start analysing the current scenarios on scale, you may realise that it may not seem so far fetched as it does.

Q19. Globally and in India, where is agri-tech and where do you see it heading?

I say this to everyone who puts up a similar question. Keep a track of Agtech for next 4-5 years.

A lot is going to change. Technology and solutions will revolutionise the way farming is done. We will move towards a more efficient and sustainable ways of farming. Farmers will grow better, sell better and we will eat better. We will see technology enabled, one acre profitable farm models (something I am obsessed with currently) through which a farmer family of 4-5 will be able to live a dignified life. Given that we in India have more than 50+ million people directly/indirectly, imagine the impact Agtech will make at scale. It will revolutionise this country more than ecommerce did or more than metro-mobility tech did.

In fact I am waiting for a day when Agtech will help move the needle of the GDP that comes from Agriculture. And trust me that day may not be very far.

Interview Parts: Part 1 // Part 2

Make sure you’re subscribed to the AI Boss newsletter to keep up with the latest in AI developments across the world!

We, at NextBigWhat, are attempting to drive forward the conversation around Artificial Intelligence in India beyond headlines that scream for attention but don’t offer much, buzzwords that make your head buzz after a while, and countless utopian tales that vie for your eyes only to leave you with a scratching head and a lingering itch to actually get to the meat of the matter.

In this spirit, the #AIBoss series intends to educate, inform and elevate our readers understanding of the subject as well as the space. If you’re part of the leadership team of an AI based product, from enterprise or startup, willing to share your AI learnings – then reach out to us and become part of this initiative (you can apply using this form).

Here is the second of three parts of our interview with Aakrit Vaish (you can find the first part here, and the third part here), CEO & Co-Founder of conversational AI platform Haptik. Aakrit co-founded Haptik in 2013 as a unique mobile assistant platform before it transitioned into conversational AI and is now one of India’s foremost platforms in that space.

Q 8. A confluence of technological developments are now helping voice become a serious contender to text – how do you see this playing out and what is Haptik doing to ready itself for this?

As mentioned in the previous question, Voice is going to lead the next phase of growth for the industry and subsequently Haptik in general. It just intuitively makes sense, it’s easier and takes lesser time to speak a command than type it.

We’ve built out capabilities to build voice bots, and are first large implementation goes live next week actually. Stay tuned and watch this space 🙂

Q 9. NLP in the Indian language space on a tech-level appears to be still in the rudimentary stages. Is this impression correct? And if so, what are Haptik’s plans in terms of working with Indian languages within chatbots?

Yes, you are right when you say it’s still rudimentary in terms of technology. But that’s also because the problem is so hard to crack. Most Indian speech is a combination of multiple languages and not just one. The concept of “Hinglish” is unique to India and there is no set data set available for it.

75% of our R&D efforts are going only towards getting vernacular languages right. We are going to be spending a lot of time and investment here, and expect at least 5-10 live implementations across languages from us in 2019.

Q 10. How does Haptik intend to stand out in the increasingly crowded enterprise chatbot market?

Focus on few key things:

1. Largest base of consumer data – The fact that we also have a direct to consumer product means that we get real conversational data from end users, which helps train the models continuously over time. This just continues making our systems better, and always ahead of any other competition that will depend on third parties for this.
2. Dedicated use cases – We will only focus on end consumer engagement and not touch internal operation areas like HR, IT helpdesk, etc. For our focus use cases, we will build out products to solve end to end business problems which could go beyond the scope of simple chatbots.
3. Types of clients – We have a set target customer segment that is very refined and precise. We will only go after those and not touch anything outside.
4. One stop shop – Continue to build and have all capabilities in house, such that our customers don’t have to go anywhere else to solve their set of problems.

Q 11. The fledgling self-service chatbot space appears ready for take off. Is Haptik planning to launch a DIY platform? And where do you see the DIY offerings settling?

The DIY mobile app space never took off, so I have my doubts here for chatbots as well. We have no plans to have a DIY platform, though I do see there could be some niche plays here if done correctly. Smaller market and a bit of a distraction for us.

Q 12. Haptik has an AI trainer team that works on the bots – what does that training entail exactly? Could you elaborate on the process? And what kind of skills does a ‘trainer’ possess?

Yes, that is again one of our differentiators. Our AI trainers take a bunch of unorganized data sets from clients and turn them into bot workflows using Haptik tools. These data sets could be FAQ documents, training manuals, product catalogs, etc. The tool then has a number of features for them to make the conversion happen.

The skill set an AI trainer needs to have is analogous to a business analyst. They need to understand the problem the client is trying to solve, take the data, use a set of tools and accomplish the tasks.

Interview Parts: Part 1 Part 3

Make sure you’re subscribed to the AI Boss newsletter to keep up with the latest in AI developments across the world!

We, at NextBigWhat, are attempting to drive forward the conversation around Artificial Intelligence in India beyond headlines that scream for attention but don’t offer much, buzzwords that make your head buzz after a while, and countless utopian tales that vie for your eyes only to leave you with a scratching head and a lingering itch to actually get to the meat of the matter.

In this spirit, the #AIBoss series intends to educate, inform and elevate our readers understanding of the subject as well as the space. If you’re part of the leadership team of an AI based product, from enterprise or startup, willing to share your AI learnings – then reach out to us and become part of this initiative (you can apply using this form).

Here is the first of three parts of our interview with Aakrit Vaish, CEO & Co-Founder of conversational AI platform Haptik. Aakrit co-founded Haptik in 2013 as a unique mobile assistant platform before it transitioned into conversational AI and is now one of India’s foremost platforms in that space.

Q 1. For those who’re unaware, could you give us a short intro to Haptik and its history?

Haptik is a conversational AI company, started in August 2013. Here is an exclusive peak into our journey, never shared before on any public platform 🙂

Q 2. With the future in mind, how much of your resources and energy are you devoting to the B2C Haptik platforms as opposed to the Enterprise platform and why?

Most of our energy is on the enterprise platform as that’s where we see the greater need today. Automating customer support for brands through chat has proven out to be an ROI positive business case, and there is a huge addressable market.

The B2C app continues to remain an important showcase product for us, but currently is on autopilot. We will pick that back up once we find a better distribution mechanism for it.

Q 3. What were the initial challenges you encountered when entering the enterprise chatbot space? What was the evangelising process like?

Yes, the biggest challenge was actually evangelism and education. All enterprises wanted chatbots, but had no idea what business problem they wanted to use them to solve. Most of my job even today is just thought leadership about what bots can do, and more importantly cannot do.

This also lead to other challenges such as price discovery, no one was aware what is the right price which caused sales cycles to be terribly long due to unnecessary negotiations.

Q 4. Working across sectors in the enterprise space, were there any unique obstacles Haptik encountered and overcame which helped elevate the platform as a whole?

Yes, definitely there is a lot to learn from each sector which can then be applied to the platform as a whole.

For example in e-commerce, people’s patience levels are lowest because the typical query is for order status. So we had to ensure the platform is optimized for speed even if the answers may not be 100% accurate. Whereas in the case of say healthcare, people care more about accuracy of information so the bot needed to be spot on, even if it took 2-3 seconds extra.

This meant that we had to create custom levers for each of these parameters that enterprises can move up and down. That makes the entire platform more customizable and flexible to use.

Q 5. From my experience, users tend to feel a higher level of frustration when a support request fails to resolve via a chatbot than with a human at the other end. What kind of premium do you place on mapping the unique customer flows of each enterprise customer & how do you go about it?

Every single chatbot built on the Haptik platform is 100% customized, and all of the customization is done by our in house professional services team. This is a strong call we took early on when getting into the business, which was unpopular with a lot of our competitors.

A chatbot is only as good as the data you provide it.

It’s also as much a design problem as a technology problem. Each workflow needs to be thought through carefully to lead to the desired outcome. This is one of the best calls I think I have ever made as the CEO of this company, and we are actually doubling down on our delivery team’s efforts.

Q 6. Could you share any analytics on what percentage of requests are being routed through to live chat and how much is being resolved via the chatbot itself?

~90% are done through the bot, 10% go to live chat

However, important thing to note is that amongst the 90%, there may be about 10% or so that are answered incorrect. These are cases where clients take a conscious call that they don’t want to have live agents at all and are okay with even incorrect answers for the long tail of queries.

Q 7. Broadly, where is conversational AI at the moment and where is it heading? What are the areas open to innovation?

Could write an essay on this 🙂

At a high level, this is my favorite way to describe it:

And at Haptik, we have literally lived this cycle since 2013. The circle above represents 2019 and the next few years are going to be the market adoption phase. We see the power of voice technology coupled with languages (particularly in a market like India) lead the next phase of innovation.

Interview Parts: Part 2 Part 3

Make sure you’re subscribed to the AI Boss newsletter to keep up with the latest in AI developments across the world!