Prodigal news

LinkedIn Live recap: What is consumer finance intelligence?

Prodigal news

LinkedIn Live recap: What is consumer finance intelligence?

Prodigal news

LinkedIn Live recap: What is consumer finance intelligence?

Here’s the weird thing about being a pioneer (not including the covered wagons and bonnets).

If you’re the first to do something, whether it’s crossing the Rocky Mountains or using AI to transform the way we understand how consumers borrow and pay, no one quite knows what you’re doing.

At Prodigal, we’re not big fans of wagon-training across the unexplored West, but we are the pioneers of consumer finance intelligence.

So what the heck is consumer finance intelligence, anyway?

Good question! We brought our CEO and Co-Founder, Shantanu Gangal in for a LinkedIn Live to explain. Check out what he had to say.

What is consumer finance intelligence?

Consumer financial intelligence touches all kinds of financial decisions that people like you and I make - the classic kitchen or dining table discussions around investing, lending, borrowing, and all the choices that come with it.

What industries or teams does consumer finance include? 

Consumer finance includes everything from credit cards to auto loans to mortgages, which are the top three categories of loans that people take. Student loans has become a big category over the last couple years, and we are also seeing a great shift in how people think about savings and investment options. 

All of those things fall under the umbrella of consumer finance. 

In addition, there are bills and other utilities that we need to pay. Those impact our wallets every single day.

What problems does consumer finance intelligence solve?

Consumer finance intelligence helps financial services companies make better choices for their consumers. It tells them what are the right options for consumers for making additional purchases.

What are the options for them to pay off their current loans, get out of any kind of indebtedness, and make smarter, better choices that are presented to them at the right time, in the right medium, and personalized to the situation.

How can we analyze such a massive amount of data?

The way things have been happening for like the last 50 years has stayed at very, very low latency. While today, there's a lot of information that is available to consumers, and there's a lot of information that is available to lenders at the fingertips that we plan to integrate into this intelligence layer.

What is generative AI?

Generative AI is a class of machine learning that takes into account the inputs that are available and generates new recommendations based on those. 

It’s artificial intelligence we can use to create or generate something brand new. Something that it hasn't been seen before, but at the same time is tailored to your use case. So your payment history, your loan history is unique to you, and recommendations made by generative AI based on them are unique. 

What is an LLM?

LLMs are Large Language Models. Basically, it's gone into the entire corpus of human information and how words are strung together to create language, create information, and it understands them at a very, very fundamental level. And this new category called LLMs rely on billions and trillions of tokens. These are data points that have gone into training these things that make them aware and proficient at creating new conversations for people like you and me.

How does Prodigal use generative AI and large language models?

We help companies in financial services by giving a set of tailored inputs to people on their team. When they're serving their customers, we are listening to the calls as they're happening, we are listening to text messages that are being exchanged, and chats that are being exchanged. And we flip it around and say, ‘For this person at this point, this is what you should say.’ 

So we give very tailored recommendations using generative AI because we've understood the conversation up to that point. And not just that - we understand all conversations we've had with the customer across a variety of media, and then compose it all together, and create a new symphony on the fly.

Why is it important to use AI trained on consumer finance? Why not just use ChatGPT?

Great question. GPT is a subsection of LLMs - the latest newest toolkit in our assets, so we leverage it too. But at the end of the day it is a means to an end. 

The end is generating extremely personalized recommendations, and doing it in conjunction with the existing techniques. So Prodigal actually brings together all of these things and assembles them in a way that makes the whole, much, much greater than the sum of the parts. 

And we can do it because we have extremely personalized, tailored information. We have hundreds of millions of interactions with various consumers across the country. This gives us an edge because we are relying a lot on proprietary information that lenders have within themselves. 

We also rely on information that might have happened with a particular borrower to come back and provide recommendations that are a lot more personalized. For example, if you don't like to give your Social Security number over the phone, and that is something we have learned in the past, the next time our recommendation to your agent will be, ‘Hey, let's just get them verified some other way.’ 

That makes a world of difference. Because suddenly you're no longer feeling like you're stuck in Groundhog Day. Every interaction is delightful. Every interaction goes 15 or 20 feet forward, and you’re not stuck in place.

What’s the goal of Prodigal’s technologies?

Our ultimate goal is to use those technologies to unite fragmented credit and lending the entire ecosystem. 

The least common denominator across the industry right now, unfortunately, is FICO credit scores. They are extremely low latency, they're very, very backward-looking. And they're very de-personalized. 

By leveraging modern technologies, including LLMs, we’re able to not just look at your payment history, but also go much further and touch upon your conversation history or interaction history, and create an offering that is a lot more relevant to the one particular bill at hand. And as a result of it, the intelligence that we generate ends up being a lot more pointed.

How do end consumers benefit when their lenders use a solution like Prodigal’s?

Consumers get a lot more. 

Firstly, they don't feel like, again, they're stuck in Groundhog Day. If you have spoken with a bank previously, all of the interaction is cleanly summarized, highlighted, cross-referenced for the next conversation. And the recommendations, the triggers and the prompts that the agent on the other side of the line gets when they're speaking with you rely on all of those things. 

So the conversation feels a lot more fluid. It is truly omnichannel, it is truly built upon all of your interactions with the person up to that point. And as a result of that, using services like Prodigal makes it a far richer, far more seamless, more fluid experience.

What is one of the first places consumer finance teams can see results with an LLM?

Any kind of machine learning is incredibly powerful at driving a lot of automation and efficiency in your frontline staff. That's one. 

The second thing is it standardizes the experience across all the people who are representing you, and gives you that standardized output, so that you are assured they are not running afoul of any compliance violations, they are giving exceptional service every single time, every single agent representing your brand has superpowers at their fingertips that make them the best agent they can be.

A lot of agents are frustrated, because they want to serve the consumer, but they’re stuck for hours across the day in bureaucratic nonsense. And automation brings a lot of delight in taking care of things that, frankly, they don't want to do anyways.

What should organizations looking for a large language model for consumer finance pay attention to?

They should really understand if the model is built and tailored for their use case. Not just the core intent engine, but also are the workflows and user interfaces designed so that your team can really benefit from them?

At one end of the spectrum, a lot of the research in this space is free, freely available, it's in the public domain. But that doesn't make it usable for you. So having the power of the technology available to you at your fingertips in a way that you can use, build, develop, and deploy tomorrow makes a world of difference for financial services. 

Prodigal news