Menu
- September 21, 2024
AI Proof of Value for Call Center
JLA assessed workflows, identified AI use cases, and developed a proof-of-value demonstrating improved efficiency and reduced costs with generative AI.
Lets Discuss Your business goals
About the client
A major U.S. business process outsourcing (BPO) firm serving telecom, broadband, and fixed wireless providers.
Operates multiple call centers, delivering Tier 1 customer support.
The Challenge
01
The customer support workflow was highly complex, involving multiple accounts, tools, and interfaces.
02
High customer service agent turnover due to complex processes and costly onboarding.
02
Facing customer attrition due to quality problems and long wait times, coupled with increasing competition from other providers.
JLA Approach
- Mapped the existing end-to-end call flow and conducted a detailed analysis to understand the volume of calls by customer, issue, network tools, and regular versus escalated calls.
- Identified multiple opportunities to use generative AI to streamline the workflow, focusing on an AI chatbot for network tool automation and knowledge retrieval.
- Set up an end-to-end secure test environment—including automated authentication, cloud infrastructure, vector database, large language model, and retrieval-augmented generation (RAG).
- Developed the POV, encompassing AI model training, prompt engineering, information extraction from network tools and contact center systems, and UI/UX design.
- Created a test plan and engaged testers to evaluate the POV and provide feedback on improvements and overall impact.
- Developed a comprehensive business case with multiple revenue streams, capital and operational cost estimates, and sensitivity analysis to demonstrate ROI over five years.
Outcomes
01
Clear demonstration of value and buy-in from key stakeholders regarding AI’s ability to assist agents in resolving customer issues more efficiently and reducing call times.
02
A detailed productization plan outlining changes to existing workflows, system integrations, testing, training, and launch.
03
A secure sandbox for experimenting with new AI ideas.
Stay in the know!
- Article | 2025-02-10
- Artificial Intelligence
- Artificial Intelligence
- Article | 2025-02-07
- Artificial IntelligenceSaaS
- Artificial IntelligenceSaaS
- Zonal Architecture