1Mby1M AI Investor Forum: Shripati Acharya, Managing Partner at Priven Advisors (Part 4)

Sramana Mitra: Yes. In all our discussions last year on AI, human centric AI was front and center. AI is in enterprises, AI is in small businesses, but it’s either a human using AI as whether it’s a co-pilot or an assistant, whatever; but it’s the human making the decisions. It’s the human using the AI to do his or her job faster, more efficiently, et cetera.

So, I think from a roadmap point of view, the agent can start to automate some of these things as the human gets more comfortable with the AI’s functionality.

I think your point is absolutely perfect in that, if the human starts to get comfortable with what the AI is recommending, et cetera, and at some point, it makes sense to automate a piece of that, that’s where the easiest insertion of AI agent driven automation comes in.

Let’s just take an example for this one. I’ve been thinking about this for a long time. In financial management and wealth management, what do wealth management advisors do? What they do can be done better by AI, most likely, right? It can be done orders of magnitude better by AI because there’s a lot of granular monitoring involved and taking decisions based on signals that are coming digitally.

If in some company, insiders are selling stocks, then I need to put a stop loss on my holding, or whatever. For these kinds of decisions, if you have an AI agent monitoring your portfolio 24 hours, there is real opportunity for agentic automation, not just human co-pilot with agent automation. So I think, again, from a framework point of view where there is a 24-hour monitoring required and where 24-hour monitoring could help in making decisions, AI would do better than humans.

If it’s a time sensitive action that you need to take, that would be a very good place to insert AI agent automation that goes and does stuff.

For things that wait, things that can be reviewed, there is time lag. It is okay to have time lag. That’s where humans still control. Humans still want to review stuff and so forth. But I think there are use cases where 24-hour monitoring and time sensitivity are important, and that’s where I think agentic AI should do very well.

Shripati Acharya: Absolutely. I think it’s a question of the comfort we have with the tools that we are using. The greater the comfort, the more we’ll be able to delegate to the agents down below.

Sramana Mitra: The other observation I have is for selling agents. If agents are being built at the top layer or the application layer, but there is a lot of integration involved throughout the workflow. Let’s say this is happening inside of enterprises; there are so many enterprise systems already within the enterprise, right? There’s the whole SAP stack, there’s the whole Oracle stack, there is the Salesforce stack, there’s ServiceNow, Workday. There’re so many enterprise stacks, and then there’re all the security stacks.

Where do you think agents are going to come from? You can’t really hook up agents from thin air. They have to be from within those systems or work within those systems.

This segment is part 4 in the series : 1Mby1M AI Investor Forum: Shripati Acharya, Managing Partner at Priven Advisors
1 2 3 4 5