Automated Virtual Agent
Objective
Enhance customer engagement and reduce operational costs by shifting the focus from traditional call center interactions to a more efficient, automated system. By implementing a conversational AI (CAI) solution, they aimed to provide round-the-clock support for common inquiries, particularly concerning their customer loyalty program, thereby decreasing the volume of direct calls and improving overall service efficiency.
Approach
1. Conduct comprehensive research to identify common customer questions and concerns regarding the loyalty program, which will guide the key functionalities of the CAI.
2. Develop and test initial prototypes of the CAI with real users, making iterative adjustments based on feedback to ensure usability and effectiveness.
3. Integrate the CAI with existing systems, ensuring it can access necessary data and perform tasks efficiently. Continuously analyze performance data to optimize responses and functionalities.
4. Regularly train and update the AI using real interaction data to enhance its responses and adapt to evolving customer needs and technology, ensuring ongoing relevance and effectiveness.
Team Structure
Innovation & Design Team Roles:
*One person may perform multiple roles.
Program Navigator, Digital Strategist, UX Researcher, Conversation AI Designer, Client, and Client Partners.
Tasks:
Research, ideate, and design proof of concept (POC) and draft minimum viable product (MVP).
Development & Delivery Team:
Project Manager, Platform Architects, Developers, Linguists and Testers.
Tasks:
Deliver, integrate, and deploy the POC and MVP. Provide support for further development.
The Solution Summary
Our team reviewed several natural language speaking platforms to evaluate and assign the appropriate technology stack. We evaluated the client’s business objectives and the capabilities of the various platforms, ultimately moving forward with Goggle Dialogflow. We chose that option mainly for its LivePerson chat platform, which allows issues to be escalated from automated to an employee when necessary.
We created conversation flows, mapping them out for the client and handling the chatbot training. As the number of intents grew, the complexity increased. We incorporated a series of best practices from our experience, utilizing lessons from previous engagements, and testing repeatedly to ensure an acceptable accuracy rate. From there, we created a setup where the intelligent assistant was installed on the airline’s website to automatically handle conversations from customers, both simple and complex, dealing with the airlines’ loyalty program, escalating issues when needed.
