Fin

What is Fin?

Fin is Intercom's highest-performing AI agent designed specifically for customer service. It resolves complex customer queries with accurate, conversational answers based solely on a company's support content. Rather than being a simple chatbot, Fin operates through a continuous improvement loop called the Fin Flywheel with four stages: Train (configure with procedures, knowledge, and policies), Test (simulate conversations before launch), Deploy (launch across all channels), and Analyze (improve performance with insights). Fin supports conversations in 45+ languages, achieves an average resolution rate of 66%, and can handle both informational queries and complex tasks like failed payments, delayed deliveries, and subscription changes.

Key Features

• Fin Flywheel Continuous Improvement

  Four-stage loop (Train, Test, Deploy, Analyze) that enables continuous performance optimization without engineering resources, allowing teams to configure, test, and improve Fin systematically.

• Omnichannel Deployment

  Deploy Fin across voice, email, live chat, WhatsApp, SMS, Slack, Discord, and social media for consistent support wherever customers reach out.

• Complex Query Resolution

  Handles advanced tasks beyond FAQs including accessing personal data, reading/writing to backend systems, following company policies, troubleshooting, and resolving issues like failed payments and subscription changes.

• Multi-source Generative Answers

  Builds answers using only the most relevant information from multiple knowledge sources including Help Center articles, internal support content, PDFs, and webpages for complete and accurate responses.

• Procedures and Tasks Automation

  Automate complex business processes with Procedures (document-style editor with code and data connectors) and Tasks that follow specific business rules for actions like canceling orders or refunding subscriptions.

• Vision and Real-time Translation

  Fin Vision lets customers share screenshots, invoices, and error messages for faster problem-solving. Real-time translation supports 45+ languages using existing support content even when not localized.

Use Cases

• Customer Service Support : Fin for Service handles the majority of customer questions including complex ones across every channel with fast, accurate answers and 24/7 availability.
• Sales Inbound Conversations : Fin for Sales runs inbound conversations end-to-end: engages prospects, guides discovery, qualifies intelligently, and closes by booking meetings and syncing with CRM.
• Ecommerce Shopping Assistance : Fin for Ecommerce acts as an online shopping assistant making personalized recommendations, surfacing upsell opportunities, and guiding customers through checkout.
• Multi-language Global Support : Companies can provide consistent support to global customers in 45+ languages without localized content, automatically detecting and resolving issues in the customer's language.
• Support Volume Reduction : Businesses achieve average conversation resolution rates of 41% (some up to 50-80%), dramatically reducing support volume while unlocking 24/7 support and faster response times.

FAQs