This is a YouTube video titled "The AI For Banks | Agentic" containing a monologue or presentation discussing the use of AI in financial services to detect fraudulent transactions efficiently, privately, and with low latency. The speaker explains the importance of owning intelligence in enterprises, particularly in handling large volumes of transaction data for fraud prediction. The content is from March 22, 2025.
Enterprises today want to own their own intelligence, right? They don't want to call APIs. One big one is for financial services. There are millions of transactions and there's a huge amount of AI that happens underneath the hood. This AI has to actually go in and understand precisely who you are based on all your transactions, what your spending activity is like, what advertisements you are likely to respond. Well, take all of that information and be able to build a prediction on if a future transaction of yours is fraudulent or not. Now doing this effectively requires there's a few of those boxes. Number one, the model has to be extremely efficient. Number two, it has to be extremely private and number three, latency. The intersection of all three of those and providing that as a solution means that we are able to provide and train very, very powerful AI systems that can ingest all of that data that's coming in from these financial service providers can predict these types of fraudulent activities in these different cases.
The AI Agent Getting More Supplier Data in 5 Weeks Than a Team Did in 2 Years
