It is said that 2025 will be the year of AI agents. Let’s explore what this could look like with Neo. I’ll use an example of an AI agent that helps users find better deals.
Example:
• The insight is that users would benefit from savings on recurring payments, but they don’t know where they overpay and switching is a hassle;
• The solution would be to use an AI agent that would do the work for them, solving the problem by reducing friction.
How can this be productized?
• Neo can already distinguish between recurrent payments vs one-time payments;
• A test group of users can be prompted to rank their experience with select subscription products;
• For example, user X pays $100 for internet; the app prompts user to give 1-5 star for value for money. If the user gives 3 stars or less, the app would ask the user if they want the AI agent to find them a better deal;
• If the user says yes, the agent can explain that they are going to search for a better deal and ask the user for some relevant info like speed or number of GB they’re getting;
• After collecting the necessary data, AI agent can scrape the internet to analyze what is available and compare available offers to what the user currently has; then come back with better options.
• If they found a better deal, they could present it to the user. If they are advanced enough, they could ask the user if the user wants the agent to unsubscribe the user from the old plan and subscribe them to a new plan. User would confirm when necessary;
• If there is no better deal, the agent can monitor for deals and suggest when one becomes available.
• By doing that, the agents would learn what deals are available and become better at predicting user preferences and matching users to relevant products.