Sramana Mitra: There’s another factor to consider. There’s this company called Tempus that went public in June this year. I don’t know if you’re tracking them. Tempus is an AI company that focuses on the pharmaceutical industry. It has all kinds of pharmaceutical use cases and healthcare use cases. It doesn’t look like a language model-based solution. It’s an older company, obviously. It already has revenue of over $500 million.
So, not all use cases are language, LLM-based use cases. There’s also machine learning applied to big data.
Rajeev Madhavan: No, everything, including, for example, the RPA company I mentioned, obviously have integrations to models of Salesforce interactions, SAP, and other complex flows. Having said that, that integration is your models is your preview, and you have to have the expertise to do that. But when you want to then summarize everything and generate some things, you could use some of these open source or Open AI kind of tools where it’s not open source, but you’re paying for it.
But, the payment terms have been coming down so radically that I don’t think that is the issue as much as having a corpus of data and models that are custom to you and differentiate you against the nearest company that’s coming behind you. If you’re just using open source data and it’s the flow, there’ll be ten people who will understand your tool in six months after you launch and will be chasing you.
Sramana Mitra: What is the state of the union on the enterprise buyer side that you’re encountering in your portfolio?
Rajeev Madhavan: In the fourth quarter of last year and the first quarter, because we have all these AI applications, I could see enterprises getting a little skirmish about buying large model driven solutions because the fear of all of the things that I discussed about coming into play.
I think that has shifted by the second quarter. Everybody’s an AI company now, right? Just like you said about Tempus, five years ago they were a pharmaceutical software company. Yes, they have incorporated AI into it. EDA companies are claiming that. Everybody’s claiming that.
So every Fortune 500 has decided to open up the purse and spend some money in those spaces – money is specifically allocated for projects that use AI now. If you are CTO at these companies and if you can’t talk the lingo and you can’t actually start taking the money and making some contributions into your products with it, you are seen as too old school. So most of them are making that transition. So, money has opened up. It hasn’t opened up into a flood, but I suspect that’ll happen towards the fourth quarter of this year and beyond because everybody wants to be the cool and say, “I’m using LLMs to do X, Y, Z.” You’re going to hear every old school company now having an AI element of a component into its product.
Sramana Mitra: What’s happening right now, where is the money opening up? Is it going more into the services companies or into the services companies that are implementing certain AI workflows? Or are they actually buying AI products from the startups?
Rajeev Madhavan: So service companies are working with a lot of startups much earlier because they will be the first one to get impacted, frankly speaking. If you are a service company, which has been providing services in, say, using automation anywhere or using SAP, 80% of your flows are going to get automated via what is incoming in tools.
Gone are the days where you could start a service company, have 50,000 people in India, and you’re charging by the number of hours, right? Because customers are going to see efficiency coming together and that will radically alter the service industry.
Service industry is going to get a lot more automated. The top service companies have all realized this and have all started spending faster. It is working with pretty much every application company that I have. Either they’re trying to invest or trying to work with them. It’s great to see that because they know that they need to make that automation layer a part of their story and their culture.
Some of them will take time to pivot because going away from an hourly inefficient flow versus an hourly efficient solution-based pricing model is a cultural shift that needs to happen in the service industry, and it will happen.
But having said that, it’s also happening in places like healthcare. Grudgingly, hospitals are realizing the changes that it can bring about to their operating procedures. It’s grudgingly because they typically are taught to treat data as a religious thing that you just can’t go around. Those restrictions have to be part of the tools provided by every good company targeting healthcare, but they need to check it. Who would be the first one to buy these tools is an important question for many hospitals. They don’t want to be the first one. So you’re going to take a lot longer in some of those industries where caution will reign rampant.
This segment is part 3 in the series : 1Mby1M Virtual Accelerator AI Investor Forum: With Rajeev Madhavan, Clear Ventures
1 2 3 4