How does
AOL
Use AI?
Achieved a 64% automated case resolution rate, improving operational efficiency.
Project Overview
AOL automates customer support by deploying a conversational AI agent that resolves billing inquiries, payment updates, and general questions for its 37 million monthly active users.
Layman's Explanation
AOL has millions of loyal users who sometimes need help with their accounts. Instead of making them wait for a human, AOL deployed a smart AI assistant from Sierra.ai. This digital helper can instantly answer questions about bills or update payment info, solving most problems on its own so human agents can focus on the really tricky stuff
Details
To modernize support for its 37 million monthly active users, AOL partnered with Sierra.ai to deploy a sophisticated conversational AI agent. The goal was to enhance efficiency while preserving the brand's reputation for high-touch, empathetic service. The AI agent was integrated into AOL's existing support channels to handle a wide range of common customer needs, maintaining a familiar experience for its loyal user base.
Powered by Sierra's AgentOS platform, the AI agent uses a "constellation" of large language models from providers like OpenAI and Anthropic to understand and generate personalized, human-like responses. Key use cases include explaining billing charges, negotiating waivers, updating credit card information, and answering general account questions. The agent is capable of performing direct actions by integrating with AOL's internal systems, allowing it to resolve issues without human intervention.
The implementation proved highly effective, with the AI agent successfully resolving 64% of cases automatically. This automation freed AOL's human support team to concentrate on more complex and sensitive customer issues that require a personal touch, successfully balancing efficiency gains with a continued commitment to quality service.
Analogy
It's like giving AOL's customer service team a fleet of brilliant, tireless digital apprentices. These apprentices handle all the routine paperwork and common questions with perfect accuracy, freeing up the master craftspeople, the human agents, to solve the unique, complex problems that require a human touch.
Machine Learning Techniques Used
- Sierra AI Agent Platform
- Natural Language Processing (NLP): for understanding the nuances of customer queries across chat, email, and voice channels.
- Retrieval-Augmented Generation (RAG): for pulling real-time, specific account information from AOL's knowledge bases to provide accurate, contextual answers.
- Sentiment Analysis: for detecting customer emotion and tone, allowing the agent to offer more empathetic responses or escalate to a human when needed.
- Ensemble Learning: (Assumption: Sierra's "constellation of models" approach, which selects the best LLM for a given task, functions as a form of ensemble method to improve reliability and performance).
More Use Cases in
Media & Entertainment
Novelty Justification
Large-scale AI chat for support is now mainstream, with most enterprises deploying similar solutions.
Project Estimates
