At Q2’s 2025 Dev Days, we gave attendees a first look at what the future of conversational banking could look like. Our demo of an AI assistant for account holders and customer service representatives isn’t just an experiment in automation. It’s a proof of concept for how natural, intelligent dialogue can help both account holders and customer or member service representatives work faster, smarter, and more securely.
Built on Q2’s Caliper API layer and powered by model context protocol (MCP), the assistant can pull real account data, provide recommendations, and help account holders resolve issues, all through intuitive conversation.
Why it matters
Today’s account holders expect more than self-service. They expect intelligent service. The assistant concept shows how AI can make digital banking more human: It understands intent, provides contextual answers, and even helps protect account holders with the right compliance guardrails in place.
And for financial institution employees, this same capability can act as a powerful job tool, summarizing unread messages, retrieving account details, and even drafting account holder responses automatically. It’s about transforming how people interact with their financial institution, both on the front line and behind the scenes.
How it works
When an account holder interacts naturally with the assistant with questions like, “Show me my accounts,” the AI draws live data from Q2’s digital banking environment through the Caliper API layer, Q2’s secure gateway for financial data and actions. Caliper manages every data request, ensuring the assistant can access only what it’s authorized to—balances, transactions, messages, or entitlements—while maintaining strict privacy and compliance boundaries.
Each prompt triggers a sequence of orchestrated API calls that work together in real time. For example, when an account holder asks the AI to “Group my recent transactions by category,” the assistant dynamically invokes APIs to fetch and classify transaction data, formats it for display, and presents an interactive summary back to the user—all within milliseconds.
This orchestration happens inside Q2’s MCP framework, which defines exactly how large language models (LLMs) interact with the financial institution’s systems. The MCP ensures that the AI can generate contextual and accurate responses without ever accessing data or actions beyond its defined scope. It’s what makes it possible to combine conversational intelligence with financial-grade security and compliance.
On the back-office side, that same MCP foundation powers a Console assistant, where a customer service representative can ask, “What did I miss since I last logged in?” and receive a prioritized summary of unread messages, open disputes, and account activity.
Behind every one of these interactions is a network of intelligent, tightly governed systems—LLM reasoning and API orchestration, all directed by Q2’s MCP layer—working together to make conversational banking real, compliant, and scalable.
See it in action
The bigger picture
This prototype is an early step toward what Q2 calls an agentic AI ecosystem, where digital assistants act with the right context, boundaries, and voice to represent each financial institution’s brand and mission.
By wrapping Q2’s APIs in the MCP, we’re defining how LLMs can safely access, interpret, and act on financial institution data. It’s a vision that goes beyond a chatbot, moving toward copilots, agents, and assistants that will soon power the next generation of digital banking experiences.
Ready to explore more AI innovation?
This is just one of the forward-looking AI developments unveiled at Q2 Dev Days 2025. Each represents a different way we’re applying AI to make digital banking smarter, faster, and more human. Read more about it.