Chatbots in Banking: A Big and Bright Future
The technological expansion and the contemporary market forces industries to evolve and update their ways. Banking is no exception and is one of the most developed sectors and avid for technological solutions. In this article, we show you what the digital transformation can do for the Banking sector, the visible benefits of those who have already bet, and some future use cases of chatbots for banks.
The first thing we do when we wake up is looking at the phone. Many find it a terrible habit, others find it necessary. However, it’s still a consequence of the times in which we live.
We have the need to connect with the world, and the fastest way is through our phones.
According to Eurostat, after viewing emails, replying to messages, and keeping up to date with the news, access to Internet Banking is at the top of activities. In addition, more than half of the European population (58%) uses online banking platforms.
Putting this into perspective and taking into account the receptivity of these innovations, it’s not so strange that banking wants to evolve at a technological level and provide a better experience to its customers in this sense.
Internet Banking and AI Chatbots
Before we were facing a global pandemic, most banks had already online platforms and non-presential customer service.
With the need to stay at home, the digital channels that once had regular affluence suffered a turnaround. The use of these means increased exponentially and companies had to quickly adopt solutions that improved and made all processes more efficient.
It was then that many financial institutions began to look for solutions with Artificial Intelligence. Based on remote interviews with banking executives, 77% states that the use of AI will be the differentiator between “the winners and the losers.”
3 Use Cases of Future Chatbots in Banking
Technological trends are changing every day, and the needs of users and companies are also changing.
“Enough” doesn’t make it anymore. We need more. It’s crucial, and customers demand not going to the branches of their banks to resolve issues of their day-to-day, which would quickly be addressed through a chatbot.
The following use cases may seem distant to you, but with the launch of PSD2 (Payment Services Directive), believe me, they’re closer than you think.
1) Opening Bank Accounts
“I want to open an account!”
This simple request is the beginning of a process that would tend to be very complicated, time-consuming, and full of unnecessary interactions.
Nowadays, to open a bank account, the best option is still to go to a branch and deal with the whole process in person. It’s true that some banks already allow you to open an account remotely. However, it still forces the user to take different steps and processes separately and always with the help of an employee.
The need to involve a human assistant often occurs at the request of regulators, but this will be another field that will undoubtedly evolve in the future.
With the digital opening of accounts, through a chatbot or other conversational channels, this process can be carried out fully automatically. However, it doesn’t dispense with human intervention whenever necessary, to offer the best possible customer experience.
“To proceed with the opening of your account, we need you to send us a proof of residence.”
When it’s necessary to submit documentation, such as an identification document or proof of address, these documents are first processed by an automatic system that not only recognizes the paper and the various text fields but also verifies whether they’re reliable or not.
“We’re almost done! We need to make a brief video call with you to validate your data. Can we proceed?”
If there is a need to make a video or phone call with a bank assistant, the chatbot itself will do it. This way, the calls occur within the chat itself, and the customer doesn’t have to change the channel.
This possibility of contacts within the chatbot significantly reduces the abandonment rate in the middle of the process.
“It follows the account opening document for your validation and signature.”
Finally, the customer digitally signs the document, and without any problems, concludes the process.
Robust and increasingly evolved authentication systems support all these steps, and we are far more secure than current methods.
Biometric Authentication and Digital Signature
There are several ways of authentication, depending on the customer’s preference and the desired degree of security.
Within the numerous forms of validation, we highlight the following. These can coincide and be in the different phases of any procedure.
- Digital signature – there’s a cross of information between the signature done and the signature of the identification document provided by the client.
- Biometric data, such as fingerprint, iris, facial or voice recognition. This data is easily cross-checked with the customer’s data, without additional actions.
2) Transactions and Personal Finance Management
“I want to transfer 10 euros to Maria Rodrigues and cancel the debit on the 28th.”
Clients can also perform operations and manage their finances. All this through chatbot, written or by voice commands.
Automatic speech recognition technologies such as Speech-to-Text (STT) help in the processing of voice requests. Additionally, it involves customer information present in their profile, namely their contacts or frequent operations (such as transfers).
With this type of solution, Banks become more and more involved in their clients’ daily routines. This way, they become a real helper in personal finance management.
“You’ve just received your salary! Can I already pass 10% of the amount to your savings account? Remember the old maxim: “Don’t save what’s left after you spend, spend what’s left after you save” Warren Buffet.”
The chances in this field are wide and the imagination is the limit. Whatever use cases your Bank wants to implement, Visor.ai team will be fully available to make it a reality.
3) Credit Requests
Another occasion when it’s necessary to spend a lot of time going to branches is in credit requests or similar. These are lengthy processes involving many paper signatures and documentation deliveries. However, this is a thing of the past.
From now on, it will be possible to simulate and define the credit you want, submit the necessary documentation, and sign the contract without leaving home (or whatever the place).
Imagine the process of buying a house. Those who have been through it know that it can be quite frustrating, especially when there’s the need to resort to home credit.
“I want to buy a house, and I want to know if your bank can lend me the money I need.”
Then you start the process of applying for a home loan.
“Surely! Can you give me more information about the house you want to buy? To make it easier, send me the link to the real estate agency”.
By sending this link, it can automatically collect the relevant data, such as:
- Location of the house – from which it’s possible to access a database with the average values per square meter in that area;
- Size of the property, characteristics, and state of the property – to facilitate comparative analysis with other similar properties;
- Value requested by the seller – to take into account the bank valuation that will be processed later.
“The location looks great! And what loan amount would you need?”
“In the meantime, I already made an offer for a lower amount than accepted, and the current owner agreed. I’ll need 200,000 € in the financing, and the total purchase amount will be 300,000 €.”
“What a good deal! Could you please provide us with your last payslips as well as your tax return so that we can analyze the operation?”
As with the digital account opening, all you have to do is drag these documents into the conversation or send them through another channel supported by the Bank. The documentation is then attached to the operation and processed both automatically.
“Taking into account the data you gave us, the Bank can ensure your financing! However, only after the Bank has made a valuation of the property, can we issue a final decision. Can we proceed with the valuation?”
Once the property evaluation report is ready, the Bank notifies the client to proceed with the process.
“Once we gather all the elements, you’ll have the Bank’s approvement!”
Perspectives for the Evolution of Visor.ai Solutions
Our product development roadmap is hugely ambitious and aims to give our customers even more functionality. We aim to provide a better level of service to our users.
Therefore, our team does not stop and is already preparing new surprises that will be released soon!
Voice Channels in Banking Chatbots
Visor.ai automation solutions already have STT technologies, in our voice bots, which allow the passage of commands from voice to text. Also, user authentication based on their voice will soon be possible.
In parallel, the chance to transform a chat conversation into a VOIP call with an assistant will be a reality. As well as the automatic transcription of conversations, in order to allow a better quality control.
The natural IVR is one of the next steps too. This will allow the bot to automatically answer simple questions by voice commands. This has been one of the most requested features, especially by those in the banking sector!
We’re already preparing this launch together with a partner company specialist in the subject.
Automatic Document Processing
OCR (Automatic Image and Document Processing) and Computer Vision are innovative technologies that, applied in the indicated contexts, are a huge asset.
As the next steps, we want to allow our customers to control all this directly in our back-office!
To do this, we’ll provide a feature called “OCR Trainer.” In this feature, simply and intuitively, you can “teach” the system to recognize new documents. It will also be possible to consult and edit the documents already processed.
Banking Chatbots: More Intelligent and Less Artificial
Currently, we have one of the best Artificial Intelligence for processing the Portuguese and Spanish languages. This system allows us to automate at least 70% of our customers’ interactions. However, we want to go further.
We want our clients to offer a service of excellence, and to provide even better user experience, both for users and customer support staff. Therefore, we’re developing processes that improve our AI, such as:
- FAQ’s Hierarchy – an excellent way to disambiguate similar requests and position them at a more detailed level. In other words, when a client asks to open a Company Account he’s no longer referred to the general FAQ about Accounts, but to the specific one about “Company Accounts”.
- FAQ’s Conflicts – it is common to have similar expressions within questions with disparate answers. With this feature, it is clear what those cases are and, in a simple way, you can correct the situation.
- Autonomous Management of Intentions and Entities – Although these features are already working successfully through the “Integration Hub”, it will soon be possible to manage them fully in our back-office and without requiring any further integration or development.
Natural Language Generator
We’re working on the creation of a “Natural Language Generator (NLG)”. This generator will allow the system to suggest the creation of answers, acording human interactions that occur between the assistants and the users. Thus, it speeds up and facilitates the maintenance of the chatbot knowledge base.
Besides, we’re developing models of cognitive analysis. These models will allow decision-making based on predetermined rules by our customers.
We cannot forget that our system is all modular, and the AI engine is no exception. Because of this, we can integrate with any solution that exists in the market.
Omnichannel = Convenience
So many times, it happens to call the contact center of a company, and, to complete or verify a process, you have to send documents. Therefore, you have to send them an e-mail with the necessary documentation and repeat the whole process you had done over the phone.
Far gone are the times when the Visor.ai platform could only integrate with chat channels. With the integration of the various digital contact channels, it will no longer be necessary to keep repeating your request.
If you are on the phone and happen to have to send the documents by email, the assistant will quickly have access to them, and they’ll be right away integrated into your process.
From the moment your process enters the system until it ends, no matter by what means you contact us, all channels will be connected and will be able to answer your requests.
For contact center assistants, they’ll have an integrated back-office for the various contact channels. There they’ll be able to easily pass conversations to other departments, migrate from one channel to another, and even gain “superpowers” to help users through the new co-browsing functionality.
Co-browsing consists of the ability to use a video call within the chatbot to share your screen with the assistant who helps you.
Design: Marta Ramos; Text: Filipa Perdigão and Gonçalo Consiglieri