Three friends - two from IIT-Kharagpur and one from BITS, Pilani – had one idea when they got out of college: do something of their own. Even as they worked their campus placement jobs, the trio would brainstorm on what this “something” could be.
The year was 2012, and selling samosas in Bengaluru was one of the “fun ideas” that crossed their minds. Foodtech was still in its early days and there was promise in India’s tech capital, especially with a large north Indian population having migrated there.
“We used to discuss everything from selling samosas in Bengaluru to high-tech artificial intelligence (AI) and machine learning (ML) ideas while trying to figure out what our startup would be focused on,” says 29-year-old Angam Parashar, who graduated from IIT-Kharagpur with a mechanical engineering degree.
Connecting the dots
Two years later, in 2014, the three set up their own Gurugram-based AI startup, Parallel Dots.
“We decided to focus on tech, our core area of expertise, and the idea of selling samosas, which was fun for sure, went out of the window,” Angam says.
The name Parallel Dots has an interesting origin – and a Steve Jobs connection. In 2005, the Apple Founder had given a speech at Stanford University where he spoke about “connecting the dots” in one’s life and the importance of “doing what you love”.
This speech had a profound effect on Angam, his classmate Ankit Narayan Singh, and their common friend, BITS Pilani graduate Mukhtabh Mayank Srivastava.
“While we definitely wanted to do what we love, we could not quite connect the dots,” recalls Angam, with a big smile on his face. It was then that his elder brother Deepak came up with the name Parallel Dots, since they were “unable to connect the dots”.
Angam met Mukhtabh at his first job out of IIT. Both were placed in US-headquartered tech and analytics company Opera Solutions’ Noida office and soon became thick friends. In the meantime, Ankit got placed in Australian mining firm Rio Tinto. He would visit India often, and the three soon became close friends.
Angam says, “From mining to data mining, Ankit’s journey is quite something.”
In under two years into their jobs, they decided to quit and start Parallel Dots.
The initial capital came from personal savings and an investment of Rs 10 lakh from Times Internet’s startup accelerator TLabs.
Back to the drawing board
The trio started out in content publishing, where their AI would pull out related and relevant content from the publisher’s archives and display them alongside the latest content. But that didn’t work too well. In 2016, they shut operations. “It wasn’t really working out. While we were generating revenue, it was not scaling,” Angam says.
The trio took a break for three months, to reassess, and go back to the drawing board. “It was a healthy mix of taking time off and brainstorming. We were constantly looking for other ideas,” he reminisces.
In June 2017, they incorporated a new company with the same name. Later that year they would raise $1.5 million from investment firm Multipoint Capital. But Angam candidly admits, “We could not decide on our primary product. We wanted to build a horizontal AI platform, meaning we wanted it to be very broad in nature, something we could sell to enterprises. But we knew we had to narrow it down in six months. That’s when we started getting requests from consumer brands. And our main product was born.”
Their product - karna.ai – became the primary business and source of revenue. Karna is Sanskrit means “to listen”. Angam further explains:
“The pitch was simple: we are building an AI platform for consumer brands and retailers across the world to collect, analyse, and visualise retail data.”
Powering analysis with AI
The startup’s clientele today includes top FMCG and consumer product goods (CPG) companies in India, US, Europe, and Japan. It also caters to big market research agencies in these geographies. The team uses its AI on image, video, and text to help these companies devise their retail strategy.
Karna.ai does retail shelf image analysis by running its AI to monitor and analyse things such as competitor shelf space, product variants in the category, the kind of marketing/ promotional offers on display in the category such as ‘but one get one free’ etc. It helps track whether the retailer is accurately executing what was commissioned by the consumer goods company in the form of offers, display etc. Angam cites an example to explain the functioning.
“Let’s say Coca-Cola has given a mom-and-pop store a fridge with its branding. The shopkeeper is only supposed to keep Coke’s range of products, but sometimes they keep competitor products like Pepsi. Our AI tracks such discrepancies and informs the company.”
On the video front, Angam explains, “Brands do eye tracking studies where an executive wearing special glasses tracks and records consumer behaviour inside a store. Manual effort is involved when it comes to analysing this video or data. We help companies decode this information through AI.”
The startup also performs text analysis, helping companies identify what is being criticised and what is being appreciated about their products by running its AI through large data sets of text reviews by customers.
Parallel Dots is currently scoping out the market to work with retail chains in domestic and western markets through pilot projects. “The information and analysis we provide are also important for the retailer as they need to know if they are making the best use of a given shelf,” Angam says. The Gurugram company also offers a “bunch of APIs” for building customised software.
The way ahead
Last week, Parallel Dots got $50,000 as the winning prize for their AI-enabled analytic algorithms in the Tech4Future challenge. The competition, which saw around 350 deep tech startups from India apply, was put together by Japanese conglomerate SoftBank Group Corp. and Invest India. The focus of the initiative was to support top Indian startups that have developed a global product in the fields of AI, machine learning, face recognition, and cyber security.
On how they plan to spend the money, Angam says, “We will invest in sales. Our biggest priority is growth and we will invest in technology.” The prize money aside, Parallel Dots will also receive incubation from SoftBank.
At present, the company is 45-member strong with a vision to grow to 65 by the year end. It has a staff mix of 50 percent in engineering, 25 percent in business, and the rest in admin and HR.
It takes on companies such as San Francisco-based alloy.ai, Istanbul-based Vispera, and Paris-based Planorama, all of which offer AI-based solutions for the global retail industry.
But Parallel Dots is clear that it has a differentiator. “Our tech sets us apart because it is scalable. Our AI can work with smaller data sets and produce fast results.”
Angam is reticent about sharing absolute numbers but says the company’s revenue has grown 5x in the last 12 months. “In the calendar year 2019 we expect another 2x growth or more,” he says. In 18-24 months, the startup wants to consider establishing a base in the US. “We hope to break even in 2019,” he says.