Overview
- Major retail and commercial bank
- Assets: $900B+
- Branches: 500+
Challenge
A leading bank aimed to transform slow, complex data access into a driver of rapid, informed decision-making. Traditional static reports and technical query requirements limited data access to a few, excluding most employees from leveraging critical insights.
Solution
Using NayaOne’s sandbox, the bank partnered with a leading AI vendor to rapidly build and refine a conversational AI system. The sandbox provided access to multiple large language models (LLMs) and cloud-based tools, enabling fast iteration without procurement delays. The PoC focused on optimising AI performance, refining user interactions, and validating scalability through real-world query testing.
Results
- 4 LLMs evaluated within 2 weeks, accelerating development.
- 5x increase in concurrent query support, proving scalability.
- 95% query accuracy for team-specific questions, empowering non-technical users.
- Foundation established for enterprise-wide conversational AI deployment.
- Chief Digital Officer (CDO)
- Business Intelligence Teams
- Customer Support Teams
A range of promising large language models, cloud-based data services, and specialised conversational AI analysis tools within the NayaOne Marketplace.
Breaking Down Barriers to Data-Driven Decision Making
Aiming to transform sluggish data access into a catalyst for rapid, informed decisions, a leading financial institution set out to provide instant, actionable insights to its non-technical workforce. The goal was clear: empower employees across departments to make smarter decisions faster, without needing to rely on technical teams or lengthy reporting cycles.
However, this ambition was constrained by the static nature of traditional business intelligence tools and the complexity of data queries, which restricted meaningful access to a small group of technically skilled users. As a result, most employees were left navigating outdated reports or waiting for ad hoc support, slowing down decision-making and limiting the organisation’s ability to act with agility.
Building a Scalable, User-First Conversational AI System
To overcome these challenges, the financial institution leveraged NayaOne’s sandbox as the ideal environment to rapidly design, build, and test a sophisticated conversational AI solution. Working in close collaboration with a leading AI vendor, the organisation conducted an accelerated Proof of Concept (PoC) that focused on both technical excellence and user-centric design.
The PoC had three core objectives:
Optimise AI performance: Evaluate multiple large language models (LLMs) and supporting infrastructure to identify the best combination for a seamless, high-quality conversational experience.
Improve user interaction: Continuously refine the prototype to deliver more intuitive, contextually aware responses, ensuring a natural and frictionless user experience for non-technical employees.
Ensure robustness: Rigorously validate the system’s accuracy, usability, and scalability through real-world query simulations, stress testing its ability to meet enterprise-level demands.
By using NayaOne’s sandbox, the enterprise bypassed traditional procurement barriers, enabling rapid experimentation with a wide range of technologies and achieving development milestones that would have otherwise taken months.
Real-World Impact of the PoC
NayaOne’s solution has been transformative for the client, driving substantial improvements in operational efficiency and strategic outcomes. The proof-of-concept phase, a critical component in the fintech partnership process, was cut by an astounding 80 – 90%. This rapid acceleration enables the bank to concurrently run proof-of-concepts with multiple vendors in weeks, a significant improvement from the traditional 12- 18 month timeframe.
