[Techie Tuesday] Vikalp Sahni's journey: from creating GoIbibo’s core tech to being a part of Aarogya Setu app
Vikalp Sahni, founding member and former CTO of GoIbibo, recently bid adieu to the Gurugram-based travel company. Taking to social media platform Twitter, he announced his decision after a decade-long journey.
“The move was in the works for close to a year now, and I was in talks with Deep (Kalra) and Rajesh (Magow). GoIbibo has reached a good scale, and there are strong teams managing everything,” Vikalp says.
The reason why he quit GoIbibo was clear. “I now want to get back to building something from the ground up. I want to build a global tech product from India that we all can boast about,” he says.
Vikalp coded regularly until his last day at GoIbibo. The opportunity to code as CTO was minimal, but he continued being hands-on.
“I always picked one project to contribute with hands-on coding. Being an engineering manager doesn’t mean you don’t code or be a part of design thinking. An engineer never stops learning; it is important to be hands-on or you will become redundant,” he says.
Prior to his exit from GoIbibo, he was working on a code to collect data on the kind of aircraft flying overhead based on radio frequencies.
Incidentally, Vikalp also was one of the core architects and volunteers in building the Aarogaya Setu app, which has been developed by the Indian government to connect people with essential health services amid the fight against COVID-19.
Tryst with programming
Vikalp, who hails from a Punjabi family, grew up in states across India as his father, a banker at Union Bank of India, was routinely transferred.
In 1997, when banks started getting computerised, Vikalp’s father joined computer courses. He would often bring home books on Windows operating systems.
“I would read those books. When I was in Class 3, my father got a book in Hindi on computers - Sanganak ek Mitra. Sanganak is Hindi for computers,” Vikalp recollects. The book got him further interested in computers.
In 2000, old computers from Vikalp’s father’s office went on sale, and the family got home a 386 Windows machine.
“That is when my tryst with programming started. And it was simple…the results were there for one to see with one small line of code,” he says. By the time he passed out of Class 12 in 2002, Vikalp had already experimented with different forms of programming and code.
From NDA to engineering
However, engineering wasn’t Vikalp’s course of choice. Keen to join the armed forces, he was studying for the NDA. Serendipitously, one of his friends told him about engineering courses and classes.
Vikalp ended up appearing for both, the engineering and NDA exams. But he knew where his heart lay and didn’t bother looking at his NDA result.
In 2002, Vikalp joined the civil engineering course at NIT Silchar. He was one of the few people who had learnt programming before joining his course. In his first year, he participated in a programming contest – the only first year student to participate - and programmed a race car simulation in an action script. While others were doing their contest in C and C++, Vikalp wanted to build a graphic racing car in action script. He ended up winning a prize for his code.
Soon, he was participating in several tech and coding contests. In 2003-2004, he built systems to bring internet to the hostels.
In August 2006, Vikalp was placed at IBM in Gurugram.
“I was the only student to go to Gurugram. All my batchmates were in Bengaluru, and I wanted to go there. Fortunately, I was transferred to Bengaluru in a few months,” he says.
The shift to GoIbibo
A few months later, in 2007, Vikalp's college mate's brother Ruban Phukan, was building a vertical search engine: Bixee. The duo had been in touch, and Vikalp liked the thought of building an India-based search engine.
During this time, Vikalp was looking to do something of his own. He wanted to do something that was more impactful. When Bixee was acquired by MIH India, Naspers, Vikalp joined as one of the core engineers. The team, which was initially building social network and vertical search engines, later also started gaming - this was what Ibibo was based on.
During this time Vikalp also built CrawlX, a crawl engine that competed with Google and Facebook as a part of Ibibo. “I built One Family, a social network platform based on family trees, as India is very family-oriented country,” he says.
Soon, around early 2009 the team decided to enter into the world of transactions, as they realised that they were competing with the likes of Google and Facebook.
"We also saw a big boom in online transactions in India" says Vikalp. This was also the time when Flipkart and Ola had started up increasing the number of online users.
While many products failed, the crawling and search features – based on the need of the consumers - worked. The successes were vertical platforms on news, auto, finance, matrimony, and travel. This was further streamlined to focus deeper on travel and gave birth to Goibibo in 2009, which was founded by Ashish Kashyap, Sanjay Bhasin, Deepak Tuli, Vikalp, and Uma Shankar,
Focus on travel
“We realised early on that travel in India wouldn’t be as fragmented a market as the Europe and US,” Vikalp says. Focusing on travel, GoIbibo decided to build fast.
Ashish and Sanjay handled the business in Gurugram, while Vikalp worked on the tech. GoIbibo was built in three weeks. In 2009, it was a fast -performing desktop website but wasn’t the first in the game. MakeMyTrip, Cleartrip, and Yatra had been launched, and were strong in the market.
But GoIbibo did one thing differently. Instead of partnering with websites, it partnered with via.com, a consolidator.
Vikalp says, “We wanted to see how we could build on top of that layer; we had to give a fast and reliable platform. At that time, if you would go to a website and search for flights on MakeMyTrip or Cleartrip, you would be taken to another page that would say ‘hold on, we are fetching your flight details’. We thought why wait for the page? We integrated the API, added the data, and showed the details. We would show details of the airline that came up with the particulars first.”
The team had realised that the way to grow was by finding different hacks. Transparency and ease of usage became our key differentiators, Vikalp says.
“I would sit on one computer on Cleartrip, Sanjay on MakeMyTrip, and Ashish on Goibibo. We would test who was the quickest to complete a booking. If Goibibo failed, it meant I had to go back to the drawing board,” he reminisces.
To stay ahead of the curve, Vikalp and his team used newer languages and technology. While the world was on Java and PhP, Goibibo had decided to move to Python.
Building network effects
The core focus was a better consumer. Vikalp explains that the Goibibo team worked on one principle: fail fast and move on.
“We would code during the day and respond to customer emails at night. It had its pluses; we were the builders and anything could be fixed immediately. It also gave a completely ‘wow’ experience to the customer,” Vikalp says.
Within two to three months, the team started seeing more than 1,000 paying customers.
The team started to grow, from two to eight and soon touched 30 people. “When we started doing 4,000 transactions, we began investing in bus and other verticals. We realised network effects were very important to keep a business running. Buses could build a network and so could hotels, but flights could not,” Vikalp says.
By network effects, Vikalp means a product or service gaining additional value as more people use it. “We knew we had to build solid network effects.” By 2013, Ibibo acquired redBus.
Vikalp says these years were pivotal. The team looked more closely at hotels as well, building software for hoteliers and customers - and that is where the chain of network effects was built.
By the end of 2017, the travel industry saw one of the biggest consolidations: GoIbibo merged with MakeMyTrip. Even Deep acknowledges that MakeMyTrip was completely blindsided about GoIbibo for the first two years.
GoIbibo then focused on artificial intelligence (AI) and voice-over text. Vikalp explains they worked out integrations with WhatsApp, Facebook Messenger, Alexa, Echo, and Google Assistant.
“We started looking at search and discovery through voice and text, to make booking easier for customers,” Vikalp explains.
He says the idea was to create value for shareholders, through tech and value optimisation. In 2019, the team focused on platform projects to build value for both companies.
“We had $16 million in overall quarterly losses. There was a time it hit $16 million in a month but we finally touched $10 million in overall losses. It was sheer bad luck that COVID hit,” Vikalp says.
In March, when the COVID-19 pandemic began raging, Vikalp felt that contract tracing was going to be important. It was during this time that his colleague, Rahul Goyal, suggested the use of GPS for contract tracing.
While the duo was ideating on this, the government was focused on building an app for the same purpose - Corona Kavach.
Simultaneously, Vikalp and Rahul, with the help of developers from GoIbibo, first built the rudimentary Aarogya Setu app with GPS and soon improvised it to use Bluetooth. They built the PoC in two days and showed Deep the prototype of the app that could help in contract tracing. Deep in turn shared the idea and prototype with Niti Aayog, who facilitated the further development of Aarogya Setu.
“We knew we couldn’t work on it alone, so we went to the government and Aarogya Setu became a completely volunteer-driven government initiative,” Vikalp says. Numerous startups, including 1mg, Qualcomm, and DailyHunt, and institutions like IIT Madras and IISC were part of this.
“It was an interesting time. We worked on a volunteer basis, but the fun of shipping code every night is unforgettable,” Vikalp says.
It was around this time that Vikalp realised it was time to move on and do something of his own; he decided to leave GoIbibo. On a break at present, he is looking at blockchain and other sectors to start up in.
What does he look for in techie? “I always look at the way people approach a problem and their mindset,” he says, quoting Netflix Co-founder Reed Hastings: “The cost of brilliant jerks is way too high.”
“Techies need to be open to learning, but more importantly to unlearning as well. Whatever you were doing three years ago isn’t sensible now. Question status quo; the dynamics always change. And don’t get emotional about a particular code,” Vikalp says.
(Edited by Teja Lele Desai)