India’s AI industry is estimated to cross $1 trillion in value by the year 2030, and that will be built on enterprise adoption. Menon believes Indian companies will deploy AI thoughtfully for value creation. He believes that “fear-mongering” about AI taking away human jobs, doesn’t recognise that use of AI will create different jobs and industries. “You think about our Ava agent as an arms merchant who will use the best gun as required,” he talks about the new agentic AI workforce that Avaamo is building, detailing plans to make it relevant across different functions across industries. Edited excerpts.

Q. How do you assess India’s AI potential, numbers pegging it between $1.2 trillion to $1.5 trillion over the next few years, particularly with scope for tangible implementations for businesses and consumers?

Ram Menon: I think it’s a mixed bag. When we start talking about AI, the traditional discussion borders around LLMs and generative AI tool sets. I feel that defeats a point for a country like India. There are three vectors on which AI adoption which we put together in one big monolithic mass for high purposes, depends.

The first vector is not exactly positive, but it’s going to really rock the industry that India created based on labor arbitrage. I always say, some of these office buildings and call centres are going to go empty. What Avaamo is doing, for example, is converting labor or services into software. I was just talking to a particular customer, and they have 3000 call centre agents, and now Ava can do that job. If you want to add more volume, all you need to add is a couple of servers. You don’t need to hire people, don’t need to train them, don’t need software, desks, offices and infrastructure. It’s all sitting on a data centre that Avaamo manages. Business models and multiple billion-dollar companies created in the last 30 years, are slowly going to be obliterated. Inshore, offshore, outsource, it’s going to be a steady drip as the technology slowly starts replacing labor.

The second vector is how India Inc. is going to deploy AI to enhance productivity. There is significant wariness about how to incorporate that into existing business operations. All this talk about LLMs and pilots, but businesses are not going to change processes because of AI. The question is, how can I make everything faster. There are some forward-looking companies, including Wipro which is one of our customers. Wipro actually has one of the largest generative AI implementations in the world for 2,47,000 employees. It is all Avaamo. Now, their viewpoint is not just about can I remove some people, but cycle time. It takes 30 days to do billing. Now with Avaamo, I can do it in 30 minutes. India Inc. will deploy AI thoughtfully for value creation, which I think is still in its formative stage.

A third vector, which is not really discussed as much, is AI deployment that is uniquely for India. For example, a basic LLM is an amazing development. We are following it to see what we can do if we deploy it for an Indian customer. If you get a Naga dialect and you get it to converse in Marathi for example, you normalise how to do customer support in India. There are some very India-oriented things, such as two or three startups, who are delivering real-time rubber prices to the farmer, and eliminating the middle-man. That is how I see AI adoption in India.

Q. In the past few months, AI has taken significant leaps forward in terms of capabilities. How do you see these developments from OpenAI, DeepSeek, Google and everyone else define where we’re headed, and is this sort of development momentum a means to an end?

RM: There are two ways to think about the maturity of AI. Most of the developments in AI are now being compressed in about 36 months. If you look at when ChatGPT was released to now, the whole conversation of AI has been about LLMs. I believe LLM is a commodity. You think about our Ava agent as an arms merchant who will use the best gun as required. The models are changing every month. Now our customers do not ask, but they were asking 18 months ago if we were using OpenAI’s models or someone else’s. We just mix and match depending on what is correct for that particular use case. In some cases, we use multiple models to deliver the value that a customer wants.

The means to an end is particularly important for a country like India. There is no one right answer to this, and as I said, it will deploy into three vectors. What can you do for an existing industry that will be disrupted? In India, an existing industry that will be disrupted will be the BPOs whereas in the US, an industry that is being disrupted is the legal fraternity. In the US, one would hire legal analysts to read through hundreds and hundreds of cases to build a brief and you’d pay them $300 an hour. Now, Harvey AI will do that in 30 minutes, and you just buy the software. The pace of change will quicken.

Q. Tell us more about Avaamo’s vision for the Agentic AI that’s just been launched, and quite what sort of vision do you see with enterprises deploying these AI powered digital workers?

RM: I’ll try to explain how this happens, and contrast it with how I see the current conversation among technology companies. Indian technologists, IT organisations, pretty much everybody is influenced by mega vendors and their marketing around new LLMs. “I’ve got this new tool set, oh my toolset for agentic platform is does this, and I’ve got these speeds” and so on. What we do is, instead of selling a tool set to a customer, we sell the replacement of a role. The use case is not a transaction in the new world of AI, the use case is we are creating is a digital worker that displays human level intelligence to replace a function or a role. Our demo of Ava is set amidst US healthcare systems. There is a number you can call and you talk to a patient coordinator, and they tell you which doctor in that healthcare system is available, will they accept your insurance, the location and so on. There is a real person who does that job right. In large healthcare systems, they can 1000s of people who do that job right. What we have built is a human clone, or an AI clone that replicates that role. It requires some understanding of content, ability to do multiple transactions, because that’s what humans do. That’s our strategy.

Over the coming months, we will roll that out for customer support, for IT support, and for employee support in companies. Employee support in a large company such as some of our customers including Aditya Birla Group, Wipro and Air India, such as getting employees to enroll for benefits, is a humongous task. There are all kinds of emails going, you fill in a form, and it’s a painful process but you do it every year. Those roles don’t need to happen anymore if you have an agent for that. We use AI and build all those skill sets for a role because that makes it easy for the enterprise to consume, since it isn’t a disparate set of tools such as a voice model, a messaging API and so on.

Q. But what about humans losing jobs?

RM: I have a very different view on this, and let me use the metaphor that crystallised in how humanity handled other disruptions in history. When Henry Ford created the Model T, and people starting buying an automobile, everyone said that all the people who are handling horses, those who build coaches and those who build bullock carts will be out of a job. That was somewhat true. However, we fail to understand the concept of new industries that would come out of that — the construction of a highway system which I estimate created millions of jobs, the construction of gas stations accompanied by stores, and motels. I think the same metaphor applies in an industry like BPOs in India. If we were to eliminate a call centre, there are a host of new jobs that are going to be created in this industry. The rise of AI has created a whole industry around labelling data, and is one of the fastest growing industries. Any LLM requires data that has to be labeled very well, and that requires millions of man hours. It’s just one example.

It is yet not clear to me what new industries will be created, because the new industries that are created because of AI in the US might be very different from the industries that will be created in India, because workplaces are different. We are a nine-year-old company here in Bangalore, building AI for the world. We have gone through three generations of AI, and we know there are a lot of smart people in India. I believe there’s a balance that some existing jobs will be disrupted and the number of jobs disrupted is primarily based on consumption. Our success is really to present the customer with a holistic package of AI for a human agent that they’re able to consume, and very quickly.