While most people want to save more, the vast majority have a hard time putting money aside to meet their own savings goals. In fact, a recent survey found that as many as 63 percent of Americans are unable to set aside enough money (at least $1,000) to cover unexpected or emergency expenses.

But there is no need to panic if your account balance isn’t exactly where you hoped it would be. You can now get help from an unexpected source — AI-powered bots provided by your bank. If you walk into a bank branch and ask how you can adjust your spending habits to save more, chances are you’ll get blank stares in return. If you call with the same question, you’ll likely be transferred from department to department. Eventually, you may be offered an opportunity to meet with a financial adviser who can help you build an investment portfolio.

With the vast amounts of your personal data being collected and stored by your bank — including balances, transactions, and payments — you’d think your bank would be able to offer you better and more customized guidance along your path to financial success. But meeting your financial goals is more about changing behavior — your personal spending and savings patterns — than it is about picking stock and bonds. And this attention to personal spending habits is something that is sorely missing in today’s banking relationships.

That’s where an AI-powered bot can be a game changer. Instead of speaking with a banker who just learned your name and who glanced at your balances and account history for 15 seconds before you started the conversation, you can now get the help of a bot that within 150 milliseconds can analyze your most relevant financial activity and suggest practical steps you can take to meet both short- and long-term financial objectives.

But a bot can do much more than just give advice and point you in the right direction. For most people, the real challenge is staying the course; it’s just too easy to fall back on old habits. Unlike a banker, the bot is there at all times (forget 9 to 4 banking hours!) to monitor your spending and the progress you make on your savings over time. It can remind you to cut back on spending when you stray from your plan and pop up at just the right time with tips to help you boost your savings. This might mean advising you when a large deposit is received or before you go on your usual weekend spending spree.

But the big question is, are bots ready to step up to the plate with such heavy responsibility? Many banking experts claim bots are not yet ready for prime time. Facebook revealed that over 30,000 bots have been created since it opened up its Messenger platform to third-party developers in April 2016. As expected, this bot gold rush has turned up a lot of dust with just a few nuggets. Examples of poorly designed bots are too easy to find, and many of them will make you giggle or shrug at their simplicity and lack of effectiveness.

While chatbots in financial services are new and will take time to mature, there are several skills they have to master on day one in order to be considered useful and trustworthy by customers.

Many of the bots we see today utilize simple natural language processing that seems cool at first but quickly falls apart as the conversation progresses. To carry an intelligent dialogue, the bot must be able to maintain the context of the conversation.

Here’s an example of a typical question a bot needs to be able to handle: How much did I spend on dining out in July? How about August?

At the same time, a bot has to be flexible enough to recognize that natural conversations don’t always progress linearly — the bot must be able to process an unexpected reply and adapt to changes in the course of the conversation.

For a bot to be helpful, it needs to really know you. It’s not enough to simply access real-time financial data and history, it must also be able to analyze that information to understand your financial behavior and to come up with smart recommendations based on your personal goals. That’s not a trivial task, and that’s why bots need to be purpose-built — you can’t expect a bot to know everything, but you want it to have deep knowledge of the issue you are trying to get help with.

As the bot gets to know you better, it must learn and get smarter over time. While banks are rightly guarded when it comes to letting AI run wild, machine learning is critical to helping the bot improve over time.

Ron Shevlin, director of research at Cornerstone Advisors and author of the book, Smarter Bank: Why Money Management Is More Important Than Money Movement, says financial institutions will have to deliver personalized guidance and advice through digital devices and channels if they want to gain consumers’ trust and maintain their relationships.

If he and many other banking experts have it right, chances are it will not be long before your bank’s bot steps in to help you save money and reach your financial goals.