Arup Roy, Vice President, Research, Gartner, during a conversation with Amit Singh revealed that AI start-ups are so far grabbing the limelight with their innovative solutions. He also underlined that India’s success in AI space depends on the government’s commitment to lay down clear guidelines and framework
How is AI landscape taking shape in India?
AI in India is most talked about for its various segments like robotic process automation (RPA), chatbots, machine learning (ML), image tagging/classification, and natural language processing (NLP).
From the customer perspective, the most talked about technology is RPA, which has nothing to do with AI and is a procedure-based tool. However, it is being talked about in the same breath as AI.
On the other hand, chatbots (conversational AI) are particularly becoming quite popular as a virtual agent with a lot of interest among the clients. However, most of it is a rub-off effect. Clients have seen and heard about it and are just willing to adopt the same without getting into the business case and the viability of the deployment. In many cases, there is little clarity towards the reasoning behind implementing a chat-bot solution.
Machine learning is quite a new area from the maturity perspective and there is a lot of hype; it’s still early days for machine learning. There are very specific use cases where ML has matured and the real implementation is far and few. The specific use cases where ML is gaining maturity are fraud detection for customer churn and decision making in banks, telecom, and insurance organizations; predictive maintenance in machinery set up; credit risk analysis for loan organizations; dynamic pricing; anomaly/tumor detection in healthcare; and drones and autonomous robots, trucks and cars for automobile, logistics and warehouses.
On the process side, incorporation of ML with OCR for processing of hand-written and printed forms is being used. However, it works fine on the printed forms but fails in hand-written forms due to the disparity in writing styles. Translation of voice to text and vice versa is another process affected by ML.
Overall, AI is being largely adopted by segments like BFSI followed by retail, healthcare, manufacturing, and government (in few cases).
While we haven’t sized the AI market, we expect $1 trillion worth of value-addition from AI over the next 5 years, globally. We foresee the huge proliferation of AI and ML across the industries.
In the near future, we will see a huge proliferation of AI specifically in internal communications including applying leaves, generating emails, mail approvals and HR interactions where chatbots will be highly accepted as a solution.
With the government gearing up to formulate guidelines and policies for the AI utilization in different industries, what are your expectations from the government and how do you see the overall AI ecosystem developing in India?
The role of government in proliferating AI in India is two-fold: as a consumer and as an enabler.
As a consumer, the government can utilize AI in citizen services, security, and smart cities, which will provide a significant push to the AI market. However, we are yet to see government adopting AI or ML in a big way.
As an enabler, the government can create an environment towards AI adoption through distinctive and clear policies and guidelines. So far, India has suffered from either lack of policy or lack of clarity in the policies. We expect the government to lay down clear guidelines and framework in terms of data usage, data security, handling of bias and incentivizing on the adoption of AI. That’s quite a crucial part for the success of AI in India.
On the other hand, the vendor landscape is quite active with a healthy start-up ecosystem. There are over 150 start-ups working in the AI field. Even the large technology organizations are heavily investing on AI such as TCS with its Ignio platform, Infosys with Nia, HCL with DRYiCE, and IBM with its Watson platform.
Enterprises, on the other hand, are evaluating the technology with pilot projects. However, we are still far behind the US and China in terms of adoption and maturity of the AI solutions. We need to develop the ecosystem of the technology providers.
Major challenges in the AI domain in India correspond to security, integration complexity and lack of internal skills. The enterprises are not very comfortable with the data security in the AI solutions as they are yet to prove themselves as a fail-proof system. In addition, the challenge is to find the right partner with appropriate skills in AI solution design and implementation.
What is the level of competition start-ups are giving to the large vendors in the AI space in India?
Startups are currently enjoying a great amount of limelight. In fact, many of the large enterprises like HDFC Bank and ICICI Bank are using AI solutions from start-ups. In fact, the enterprises are ready to explore this new area with small players having innovative and out of the box solutions. Moreover, a majority of the AI revenues in India are accounted for by these small start-ups.