The Customer Service Revolution: Building Fin, with Eoghan McCabe & Fergal Reid of Intercom

In this episode of The Cognitive Revolution, Nathan Labenz interviews Eoghan McCabe and Fergal Reid from Intercom about their AI customer service agent, Fin. They discuss how Intercom stays current with AI, the importance of testing in production, and the pressure to be quick to market with new AI capabilities. They also explore the real-world impact of AI on customer service, the factors driving Fin's resolution rate, and the future of AI in customer service. They touch on the challenges of deploying AI, the importance of speed in customer service, and the potential for AI to expand into other areas of the customer lifecycle. Finally, they discuss the impact of AI on Intercom's business, including their pay-per-outcome pricing model and the goal of doubling engineering output with AI assistance.

Outlines

Sign in to continue reading, translating and more.

Continue

"The Cognitive Revolution" | AI Builders, Researchers, and Live Player Analysis

Introduction to Intercom's AI Strategy and Keeping Up with AI Developments

Internal AI Team Structure, Experimentation, and Evaluation Strategies

Market Pressures, Incumbency Advantages, and the Pace of AI Adoption

Balancing Innovation with Customer Confidence and the Real-World Impact of AI on Employment

Fin's Resolution Rate, Performance Drivers, and Future Investments

Intelligence as a Bottleneck, Model Size, and Missing Capabilities in AI

Balancing Accuracy Expectations, Human Error Rates, and the Impact of Speed

Paradigm Shifts in AI Agent Design and the Future of Intercom's Product

The 99 Cents per Resolution Pricing Model: Incentives and Value

API Usage, Product Philosophy, and Feature Discovery

AI Adoption Across Functions, Engineering Efficiency, and Expanding into Adjacent Niches

The Customer Service Revolution: Building Fin, with Eoghan McCabe & Fergal Reid of Intercom

"The Cognitive Revolution" | AI Builders, Researchers, and Live Player Analysis

00:00Introduction to Intercom's AI Strategy and Keeping Up with AI Developments

Introduction to Intercom's AI Strategy and Keeping Up with AI Developments

04:48Internal AI Team Structure, Experimentation, and Evaluation Strategies

Internal AI Team Structure, Experimentation, and Evaluation Strategies

13:04Market Pressures, Incumbency Advantages, and the Pace of AI Adoption

Market Pressures, Incumbency Advantages, and the Pace of AI Adoption

21:34Balancing Innovation with Customer Confidence and the Real-World Impact of AI on Employment

Balancing Innovation with Customer Confidence and the Real-World Impact of AI on Employment

36:18Fin's Resolution Rate, Performance Drivers, and Future Investments

Fin's Resolution Rate, Performance Drivers, and Future Investments

45:05Intelligence as a Bottleneck, Model Size, and Missing Capabilities in AI

Intelligence as a Bottleneck, Model Size, and Missing Capabilities in AI

55:56Balancing Accuracy Expectations, Human Error Rates, and the Impact of Speed

Balancing Accuracy Expectations, Human Error Rates, and the Impact of Speed

1:04:53Paradigm Shifts in AI Agent Design and the Future of Intercom's Product

Paradigm Shifts in AI Agent Design and the Future of Intercom's Product

1:11:04The 99 Cents per Resolution Pricing Model: Incentives and Value

The 99 Cents per Resolution Pricing Model: Incentives and Value

1:18:51API Usage, Product Philosophy, and Feature Discovery

API Usage, Product Philosophy, and Feature Discovery

1:27:35AI Adoption Across Functions, Engineering Efficiency, and Expanding into Adjacent Niches

AI Adoption Across Functions, Engineering Efficiency, and Expanding into Adjacent Niches