[Techie Tuesday] Swapan Rajdev of Haptik, one of the largest conversational AI platforms, has one advice for techies: build technology that has a purpose
As the Co-founder and chief technology officer (CTO) of Haptik, Swapan Rajdev’s journey as an entrepreneur was filled with passion and many challenges.
From building a gaming app for iOS, which was ranked number four on the App Store in the US within a few days of launch, to building business tools for marketing and sales professionals at Radius Intelligence, a startup in Silicon Valley, Swapan is known to have built products from scratch, pushing against the tide, every time.
Today, at Haptik, a B2B solution for enterprises, which builds chatbots that companies can deploy on websites, apps, and other platforms, Swapan handles all of product and technology, and brings in, as he says, a balance between the long term and short term.
Apart from this, he is consistently working on new and innovative ways to make Haptik one of the best places to work at.
Advising young techies, Swapan says: “Build technology that has a purpose. It is great and passionate to build good technology, but that technology should have a purpose to it. I look at building code as a form of art, something that’s really going to come to life. And when it comes to life, it needs to have a purpose. You need to solve for that purpose and be obsessed about that."
Born in Delhi, Swapan and his family moved to Kuwait when he was three-years-old. Swapan’s mother was teaching children with special needs, and his father was into accounting. Needless to say, Swapan didn’t have much exposure to technology, computers, or the internet until 1994, when he was in 11th standard.
“Initially, I never got into how these things were made, and I focussed on chatting - ICQ, MIRC, and that was it. So, when in Class 11, I had to opt for an elective, and I took Biology. But within two months, I realised I just didn’t want to study the subject, and so I shifted to Computer Science,” says Swapan.
Turning from a user to a builder
Until now, Swapan hadn’t met anyone who could guide him into the world of computers. But after opting for the course at 16, Swapan knew this is what he wanted to do for the rest of his life.
“It was great to see that you write 10 lines of programme and see an output. The feeling is out of the world - a few lines of code can change something and build something. And that is when I got deeper into the workings of the computer and every nitty gritty detail,” says Swapan.
It was then that Swapan changed from being just a user to creating something on his own.
And the first thing he built was a website about himself in 2000.
This dual passion led him to opt for computer engineering course at the University of Illinois, Chicago. “I wanted a mix of hardware and software because that is where a lot of good games come around with an understanding of both,” he adds.
While the core computer engineering course focussed on a lot of hardware, Swapan also attended several software-related courses to help bring in the balance.
“At college, I coded an assembly language using registers. One of the projects I took up was to build a ‘snake’ game using an FPGA board. It is a hardware that you can code on top of, and I also built a small autonomous self-driving car. The idea was simple - leave a car at the beginning of the maze, and see if it can automatically reach the finish line,” says Swapan.
The jet setting life of a consultant
While his college life was filled with such interesting projects, Swapan had to move on to the next leg of his career.
It was around 2007, and it was the time when two things seemed to be the talk of the town - iBanking and Consultancy. For Swapan, who had just graduated, the lifestyle of a consultant seemed fascinating. “The idea of being on a plane, living in a hotel working with clients and helping them, and bringing in a mix of technical and consultancy expertise - all seemed rosy,” says Swapan.
So he joined Accenture in Chicago as a technical consultant, where he focussed on databases and data management. Here, Swapan had the task of building ETL tools for big banks and media clients to manage and store their data.
Swapan got the lifestyle he wanted. He would travel from Monday to Thursday, and return home (Chicago) by Friday.
Speaking about his days at Accenture, Swapan says: “I learnt a lot about how a large organisation works, and I understood about client interactions. It taught me the importance of communication - whether you’re an engineer or not, how you communicate is important. I also learnt to manage big data."
But after two years, Swapan felt the job wasn’t challenging enough. This was the time when the iPhone was just launched, and the ecosystem around mobile phones was gaining steam.
The world of iOS
Despite landing a job he was fascinated with, Swapan felt he wanted to do more, and so he started learning iOS development parallelly.
“I was also in touch with my college friend Aakrit Vaish, Co-founder of Haptik, who told me a lot about apps and app development. He was working at Flurry in the Bay Area,” says Swapan.
During their conversations, Swapan thought of building a game on iOS. He says:
“I was thinking of building a game for iOS. I used work in the office from 8 am to 9 pm, and from 9 pm to 4 am I would work on the game. On Friday nights, I would come back to Chicago and work out of Starbucks. Because I was learning and building together, I spent six to seven months building the game. And Aakrit helped me on how to monetise the app. Nine months later, during 2010 Thanksgiving, I released the game on the App Store in the US. Within a few days of launch, the app was ranked number four on the App Store in the US."
By 2011, Swapan realised that he could not continue in consultancy and those very things that fascinated him earlier was no longer appealing, and he moved to the Bay Area on Aakrit’s advice.
Moving to the Bay Area
Swapan chose to leave a coveted job at Accenture for a startup in Silicon Valley. “With nothing in hand, I quit my job, packed a suitcase from Chicago and moved to San Francisco. Here I met different people, was networking, and realised that the people I interacted with while at Accenture were so different from the people in Bay Area,” he says.
The same year, he joined Radius Intelligence. It was a team of seven to eight people, and the founders were the first interns at Facebook. The seniors at the office happened to be Swapan’s college alumni.
“There was no product defined at that time, but what attracted me was the kind of people. They were super smart, and I still think I was the least smart person in the crowd. The technology they were building was geotagging the internet - taking every piece of content in the net and building this graph of places and people. The product on top of the technology was a business intelligence tool for sales and marketing people,” says Swapan.
Swapan explains he was also part of the mobile journey - with the iPhone and then Android coming around, he saw the smartphone journey. He explains that being part of the Bay Area was a great learning for him as he could understand about new consumer behaviours, and how they changed.
The early days of AI and ML
“You are part of an era where people are moving from desktop to mobile. It makes you realise why the product is so important - technology doesn’t matter if you can’t package it to the right product. I realised we can build the best technology, but if it doesn’t have the right experience, design, and product, it is not going to matter,” says Swapan.
He says, being a part of Radius also gave him the early exposure needed towards machine learning and artificial intelligence.
However, even while working at Radius, Swapan didn’t stop building apps. It was during this time, around 2012, that Aakrit and Swapan would sit together and discuss different ideas.
“Aakrit and I discussed apps like WhatsApp were fast catching up. So, we felt chat and conversations is what people are going to do in the future. And this led to the birth of Haptik. We thought why not build something that builds utility on top of chat, and not just peer-to-peer chatting,” says Swapan.
They realised the most conversational use case in the world is consumer support. It got them thinking of how one can connect brands, companies, and people. “SMS is more spammy, and the idea is to build a channel that is more communicative. Having already used a bit of machine learning - you can really automate it. We decided to create a communication channel for large brands,” says Swapan.
“We quit our full-time jobs to start Haptik as we realised that we had to go all in and not do side gigs. So, in 2013, we started building this when chatbots and AI weren't big, but machine learning did exist,” he adds. While Aakrit moved back to India, Swapan continued in the Bay Area.
The duo decided to launch Haptik on March 31, 2014. However, like most entrepreneurs, there were teething problems. Brands were wary of signing them as there were no consumers, and consumers were wary because there were no brands. Also, conversational data was hardly available. So, they decided to collect the data first to build the algorithms. For this, people started responding to user queries, and this helped crack the chicken and egg problem.
“I built the iOS app, and a friend built the android app. I also built the dashboard where people started chatting. While I would take up chats and coding in the nights, Aakrit was doing so in the day.”
“In 2014, we raised our first round with Kalaari Capital while I was still in San Francisco. It gave us leeway to hire more people and build the product. That is when I started building the machine learning algorithms and bring in more people in engineering,” says Swapan.
It was then they realised that Haptik was becoming more of a personal assistant. While customer support as a consumer-first app is a great use case, the challenge was that as a user you don’t have customer support problems that often, and there was limited recall. Swapan says, their data showed people asking them simple day-to-day queries on food ordering, hotels, and flight booking.
By 2015, they realised it was difficult to build a product with the team sitting away from San Francisco. Hence, Swapan moved back to India. By 2017, the whole chatbot industry started growing with Amazon Alexa, Facebook, and everyone else building on machine learning and AI.
The Haptik technology had matured as well, and now the team was working with large enterprises to help them with their customer care experience and enable conversational commerce. While Reliance now owns majority stake in the company, the team continues to build the product and grow.
“Today, I am focussed on making sure we are building the right things with high quality and great execution. When I hire people now, I also look for a cultural fit in terms of aptitude, attitude, and skills,” says Swapan.
Speaking about what he looks for in a candidate while hiring, Swapan says: “Will we all enjoy working with this person, do they raise the bar of Haptik - not just from a technical standpoint, but what is the new viewpoint they will bring in, and what is the unique thing they will bring in.”
(Edited by Megha Reddy)