What AI brings to the table

Transcript 

00:00:00 - Introduction

Before diving into a live demo, the speakers explored the broader future of internet banking. According to recent studies, by 2030, over 80% of banking transactions are expected to be driven by AI technologies. This shift reflects changing user behavior, end users increasingly prefer digital interactions over speaking with human agents or visiting branches. In fact, younger generations are moving away from traditional transactional flows altogether.

A forecast from Juniper Research estimates that banking spend on generative AI will reach $85 billion by 2030, underscoring the scale of investment and belief in AI’s potential. Another study from Max Centre revealed that AI-powered experiences can capture over 60% of user attention, boosting digital engagement and sales while reducing operational costs. By developing AI for one channel, banks can extend their reach across platforms from internet banking and mobile apps to wearables and messaging services like WhatsApp.

The session also highlighted the natural accessibility of conversational AI. Technologies like text-to-speech, voice recognition, and chat interfaces make interactions intuitive and inclusive. While early AI chatbots focused on FAQs and knowledge bases, today’s users expect more. They want to perform transactions, simulate credit scenarios, and manage their financial well-being through conversation.

ebankIT showcased its readiness for this future with a pilot demo of a fully functional AI chatbot. Built on top of the bank’s secure infrastructure, the chatbot named Kim can handle tasks such as checking account balances, making payments, simulating credit applications, and even predicting financial trends.

During the demo, the AI chatbot successfully retrieved account information and recent transactions, demonstrating seamless integration with ebankIT’s APIs and banking stack.

00:05:00 - Conversational banking in action

The live demo continued with a deeper look at how the AI-powered chatbot handles real banking tasks. After confirming there was no incoming credit, the user initiated a money transfer to a contact named Michael. Impressively, the chatbot was able to identify the correct recipient even though only the first name was mentioned, thanks to contextual awareness. The transaction was then completed through a secure deep link, using trust factor authentication without requiring manual SMS input. This design choice, redirecting users to a confirmation screen, was intentional, ensuring clarity and security for money movement actions.

Next, the user asked Kim to simulate a mortgage loan. Integrated with ebankIT's partner: ITSCREDIT, Kim could simulate various financing options. Although initially requesting salary input, Kim intelligently retrieved the necessary data from account movements, successfully completing the simulation. While applying for the loan isn’t yet available, the simulation feature already adds significant value.

The demo also showcased Kim’s ability to analyze and categorize expenses, helping users identify duplicate charges and potential savings. When asked about recent dining experiences, Kim pulled up relevant transactions and even generated an interactive chart to visualize spending habits. Users could filter categories and explore their financial behavior in a more engaging way.

The demo concluded with a showcase of voice-enabled interaction, highlighting how the AI assistant, Kim, can support users even while they’re driving. Switching to a mobile interface, the user asked Kim to check their credit card balance and remaining credit, both of which were answered promptly and naturally. When the user mentioned they were driving, Kim responded with a friendly, safety-conscious message, reinforcing the assistant’s human-like tone and contextual awareness.

00:10:00 - How ebankIT's AI works

To close the session, the team offered a behind-the-scenes look at the technology powering the AI chatbot. At its core, the assistant uses Azure OpenAI GPT-4.1, enabling it to understand complex queries, manage multi-turn conversations, and respond in a natural, conversational way. It’s also integrated with Azure AI Speech for text-to-speech and voice recognition, making it fully functional across both text and voice channels.

The AI assistant is built on a proprietary Model Context Protocol (MCP), which allows it to maintain session awareness and tailor responses to each user’s financial context. This protocol acts as a bridge between natural language input and the banking transactions executed within the ebankIT platform.

All of this runs within ebankIT’s secure architecture, backed by robust APIs that ensure compliance, security, and reliability. The AI assistant also supports multilingual interaction, responding in the same language used by the customer, making it accessible to a global user base.

This session not only illustrated ebankIT's AI technical capabilities in banking but also reinforced its role in enhancing user experience, accessibility, and operational efficiency. As AI continues to evolve, ebankIT is positioning itself at the forefront of this transformation ready to deliver smarter, more human-centric digital banking.