Quantiphi's AI-Powered Virtual Agent Achieved 95% Success Rate in Resolving Queries Streamlining Mortgage Valuation Query at the Scale and Complexity for an Indian Multinational Bank
Read how Quantiphi's solution led to ~28K total sessions, handled 150K total queries, maintained an average of 72 sessions per day, and achieved a 95% success rate.
About the Client
The client is one of the largest Indian multinational banks, headquartered in Mumbai. With 6,000 branches and 17,000 ATMs across India, it offers a wide range of financial services globally. The bank is known for its extensive service offerings, including personal banking, corporate banking, and wealth management services, serving millions of customers.
Problem Statement
The mortgage unit of a major Indian bank faced significant challenges in managing and responding to repetitive queries. The process was labor-intensive and time-consuming, with difficulties in maintaining updated information from multiple sources on a single platform. This inefficiency led to delays and increased workload for human agents. The team needed an AI-powered virtual agent to streamline query handling, reduce human effort, and ensure quick, accurate responses.
Challenges
- Labor-intensive Process: The manual handling of repetitive queries consumed significant time and resources, leading to operational inefficiencies.
- Information Management: Consolidating updated information from various source systems into a single platform was challenging, resulting in delays and inaccuracies.
- Human Effort: The reliance on human agents for answering repetitive queries reduced their availability for more complex tasks and customer interactions.
The Solution
Our Implemented Solution:
Quantiphi developed and implemented an AI-powered virtual agent to address these challenges. The virtual agent was trained on 10,500 FAQs to efficiently handle repetitive queries, providing quick and accurate responses. The solution included the creation of an Admin Panel for managing and updating responses and a Power BI dashboard to monitor the virtual agent's performance.
Key Features :
- Virtual agent trained on 10,500 FAQs: Capable of responding to or redirecting users based on simple, repetitive queries for quick resolution.
- Creation of an Admin Panel: Enables uploading, downloading, and updating responses to existing FAQs across different lines of business.
- Power BI-based dashboard: Provides a comprehensive view to gauge the performance of the virtual agent.
- Prebuilt with Domain-Specific Data Points: The solution comes prebuilt with an extensive set of domain-specific data points, ensuring rapid deployment and faster time to market. Its flexible architecture allows for the easy integration of new datasets, which can be done within 8-10 hours, leveraging Quantiphi's deep domain-specific expertise to adapt quickly to changing needs.
- Technologies Used: Dialogflow CX, BigQuery, Google Cloud Platform, Cloud Functions, Cloud Firebase, and PowerBI were utilized to develop and support the solution.
Results and Impact Created
- High Engagement:
- 28K Total Sessions: The virtual agent experienced significant user interaction, indicating strong adoption and engagement.
- 150K Total Queries: The high volume of efficiently handled queries reduced the workload on human agents.
- Daily Usage:
- 72 Average Sessions Per Day: The consistent daily engagement demonstrated the virtual agent's reliability and usefulness in day-to-day operations.
- Efficiency and Success:
- 95% Success Rate: The virtual agent achieved a high accuracy rate in resolving queries, leading to improved customer satisfaction.
- Time Savings: The virtual agent significantly reduced the time spent on answering repetitive queries, allowing human agents to focus on more complex tasks.
- Improved Information Management:
- Unified Platform: The centralization of information from multiple sources ensured that the virtual agent provided accurate and updated responses.
- Admin Panel: The panel facilitated easy updating of FAQs and responses, keeping the information relevant and current.
- Performance Monitoring:
- Power BI Dashboard: The dashboard provided valuable insights into the virtual agent's performance, enabling continuous improvement and optimization.
- 28K Total Sessions: The virtual agent experienced significant user interaction, indicating strong adoption and engagement.
- 150K Total Queries: The high volume of efficiently handled queries reduced the workload on human agents.
- 72 Average Sessions Per Day: The consistent daily engagement demonstrated the virtual agent's reliability and usefulness in day-to-day operations.
- 95% Success Rate: The virtual agent achieved a high accuracy rate in resolving queries, leading to improved customer satisfaction.
- Time Savings: The virtual agent significantly reduced the time spent on answering repetitive queries, allowing human agents to focus on more complex tasks.
- Unified Platform: The centralization of information from multiple sources ensured that the virtual agent provided accurate and updated responses.
- Admin Panel: The panel facilitated easy updating of FAQs and responses, keeping the information relevant and current.
- Power BI Dashboard: The dashboard provided valuable insights into the virtual agent's performance, enabling continuous improvement and optimization.
Quantiphi's virtual agent solution effectively addressed the bank's challenges, enhancing operational efficiency and customer satisfaction through automation and advanced AI capabilities.
