Presenter: Umakant Soni 

Moderator: Sindhuja Balaji 

These are tough times for entrepreneurs. A recent report by NASSCOM presented a rather grim picture for the Indian startup ecosystem - 90% startups are facing a decline in revenue, nearly 40% have either halted operations or are on the verge of shutting down and 70% startups have runway less than three months. 

We, at Indiaai, thought it would be prudent to hear from a seasoned expert on why this crisis could be a landmark moment for the Indian AI ecosystem, and why this is the perfect time to sharpen product, analytics and research capabilities to become an AI leader in the new world. Here are the higlights from the INDIAai360Series Webinar #2 - COVID19: Pivoting Opportunities For Startups with Umakant Soni, cofounder of AI Foundry & founding partner of pi Ventures 

The AI Journey Over A Decade

After spending several years at Wipro, I started my own AI chatbot company about 10 years back. I also spent some time in Silicon Valley learning about building ecosystems and wanted to bring this to india. Manish (Singhal) and I set up pi Ventures, which was an AI-focused fund. We were told by many that running an early-stage fund in India is tough, especially one that's focused on a specific aspect of technology. But we raised money over a period of time and invested ins some really strong AI companies. Some have even reached around 100mn in valuation. This is a very fulfilling validation of our faith in the AI ecosystem in India. It's also interesting to see that while India is usually not the place to start a deeptech AI venture, but I feel that one can grow a company here and take it global. 

Until recently, there was not a lot of traction for AI companies in India, but trends indicate that there should be a lot more activity in this area. We went to MeitY and NITI Aayog to enhance the value proposition of this sector and that India should be at the center of this trillion dollar opportunity. Unlike Canada, Israel, China, we dont have an AI ecosystem builder. So we came up with the National AI strategy. Following that, the Govt of India invested in $1.5bn in the research & innovation ecosystem. Currently, I'm running a venture studio AI foundry, where I'm focused on building AI for the next 6bn users. In addition, I have the Artpark or the AI and Robotics Technology Park which is focused on research and will allow startups to really go do a deep dive into AI. These steps are necessary to provide the innovators of tomorrow a strong foundation. 

The COVID19 Effect On AI Growth Curve

The effect of COVID19 is bigger than WW2. During WW2, 4% of the world population got wiped out. Nearly $350bn was the cost of war. Even with the current economic situation, COVID is expected to cost us $3.4 trillion from a global workforce perspective. GDP growth rates across countries is going to be negative. Coming to AI, its onviously not new. It used to be called handcrafted machine learning systems. Most matching or recommendation systems you would have seen with apps in the early 10s would have been based on handcrafted ML systems. At the time, developers were just about skimming the surface of problems. If you were to visit an ecommerce site, you would have received the most product recommendations based on your preference, but there were still many issues to do with packaging, delivering which made the online shopping experience feel incomplete. This was Wave 1 in AI development, and could address about 15% of challenges around us. 

With Wave 2 came deep learning, which was was easier because data was more available, computing got cheaper and GPU became more accessible. Errors in NLP and Computer Vision were brought down to almost human levels. Additionally, one could truly mimic hearing, visuals and more. These added capabilities could solve around 30-35% of the problems in AI. Companies like Sigtuple and Niramai fell in this bracket. Like for a blood sample, Sigtuple was applying AI to pathology, essentially "mimicking pathologists" by allowing the camera to take pics of human samples, scanning it in real time and having a GPU on the device decode the results. While it was robust, it still needed human intervention to work seamlessly. 

Wave 3 is now starting to take off, which is humanlike AI with a combination of sensory inputs. It becomes more niche, more subjective. And AI could solve the problem more effectively. With AI being the core of any problem, we will see some really disruptive companies emerge in the times to come. The challenge of scaling with marketing is that scaling begins to slow the growth but with a competitive tech tech advantage like AI, it could not only accelerate growth but allow you to dominate the space. 65-70% of any problem will be driven by AI and this phase would see the birth of AI-first companies. 

COVID has acceleated this pace of change. Its tricky, it could cause a major shift much faster than we anticipated. COVID19 & AI is expected to wipe out industry players in verticals that are not led by technology at their core. 

Be it airlines, housing, offline retail, O&G, automotive, there will be widespread disruption in these sectors led by tech. This is a lot like the asteroid that hit the earth that led to dinosaurs getting wiped out, and allowing a new species to thrive and expand in numbers. This is called adaptive radiation. COVID19 is a black swan event for us, like the asteroid hitting the earth. Post COVID, new startups will emerge - they will be smaller, nimble, deep-tech focused, and deftly occupy niches left behind. We are going to see everything move online, automation of local and global actions, combination of physical and digital world and cost reduction brought about by automation & data. 

India will be the source of 'leapfrog innovation' in COVID AI age 

There are multiple examples that prove India enjoys a unique advantage to leapfrog in AI. In a matter of two months, we have harnessed the biggest contact tracing app in the world Aarogya Setu, which already has over 110mn users in India. Aadhaar is the world's largest citizen database, and is also powered by some of the most advanced technologies today. With Aadhaar as a foundation, several interesting innovations are taking place. We have data, a young adoption-centric population - all these factors make India a prime candidate for some of the best AI innovations. India is often called the "supermarket of problems", which will accelerate the levels of AI adoption to solve challenges and use cases unique to India and its way of living and doing business.