AI-led Contact Center Transformation for a leading financial services provider
BFSIBusiness Impacts
$2.3M
estimated annual savings
150k
calls deflected per annum, i.e., ~20% contained calls
Improved operational control through micro-level insights at the use case level and agent level
~90%
accuracy in query recognition
Customer Key Facts
- Country : United States
- Industry : BFSI
Problem Context
The client’s existing IVR-based contact center system had complex and long IVR trees resulting in long call durations and poor customer experience. This also resulted in a high call escalation rate (75-90%) with a human-intensive and cost-ineffective system. Additionally, a lack of visibility into contact center operations due to missing visualization capabilities and analyses for Contact Center KPIs further advanced inefficiencies.
The existing IVR system delivered non-personalized experiences due to its inability to account for customers’ historic interaction information and followed the same conversation path for all customers. Also, different live agents answered user queries over different calls – resulting in an inconsistent experience. These inefficiencies resulted in higher costs per user session compared to the industry average.
Lastly, the disconnected nature of the systems meant tedious maintenance efforts for the infrastructure.
Challenges
- The existing IVR-based contact center setup was highly complex and less efficient with long IVR journeys which led to longer calls
- Higher reliance on human agents due to lack of effectiveness and automation
- Impacted experience due to toggling of keypad and voice through DTMF
- Lack of in-depth data analytics to measure and plan features and fix gaps
- Comparatively higher cost of feature turnaround time and effort
Technologies Used
GCP
DialogFlow CX
Looker
Apigee
Insights
Agent Assist
Genesys Multi Cloud
Solution
Phase 1: During the first year of engagement, Quantiphi built:
- A virtual agent leveraging Dialogflow CX and Agent Assist integrated with the existing Genesys IVR system. The CCAI system is being designed to support five use cases and is configured for live agent handover for the telephony channel.
- Business and technical dashboards to monitor and analyze the performance of the virtual agent and contact center using CCAI Insights.
Results
- IVR containment rate improved from negligible levels to ~20-40% after the first deployment
- Seamless handover to live agent without repetitive messaging
- Data-driven decisions to plan feature rollouts
- Ability to measure the impact of new automation and feature rollouts
- Analysis of user behavior to plan fixes and improvements
