Fin 2

The first AI agent that delivers human-quality service

Fin depicted as an energy Orb

The only AI agent with
human capabilities

Use Fin with Intercom or your existing platform—no migration required

Product UI for Fin working within Zendesk

Get the best AI agent, and keep your team on Zendesk

Fin AI Agent can instantly resolve 50% of your support tickets, emails and messages on Zendesk. Fin delivers superior performance, answers more complex questions, and is easier to set up—creating a better experience for your team and your customers.

How Fin works on Zendesk
  • Set up in under 15 minutes
  • Handles tickets, emails, live chat, SMS and social
  • Learns all your support content from Zendesk Guide instantly
  • Follows your Zendesk ticket assignment rules
  • Hands off to your team directly in your Zendesk workspace
  • Continue using all your existing Zendesk automations and reports
Placeholder
Placeholder
Dental Intelligence

97%

Fin CSAT Score

Sam MillerCustomer Support Operations Manager

Sharesies

70%

Resolution rate with Fin

Ruby PictonInvestor Care Lead

Robin

The numbers speak for themselves. We’re seeing a 50% resolution rate with Fin, which is pretty amazing!

Ben PeakDirector, Technical Support

Linktree

Within six days, Fin is successfully resolving 42% of conversations. It’s truly surpassed my expectations.

Dane BurgessCustomer Support Director

 
 
 
 

Thousands of businesses have already seen transformational results

Features

Built to work like the best customer service agent

Fin 2 is the next generation of Fin AI Agent and its new features make it the most advanced in the industry.

Fin uses all of your support knowledge, implements the correct behaviors and policies, personalizes every customer experience, and delivers insights on how to improve its performance.

Fin 2 Illustration
1

Knowledge

Uses all of your knowledge to craft the right answer

Fin has complete, up-to-date knowledge of all your products and company processes. It provides customers with the deepest, most thorough answers to help resolve issues faster and more efficiently than ever before.

A screenshot of the UI showing the Multi-Source Generative Answers feature

Multi-Source Generative Answers

Fin constructs answers using only the most relevant information from multiple knowledge sources, allowing it to create more comprehensive responses, and enhancing its ability to solve complex questions.

A screenshot of the UI showing the Fin Knowledge Management feature

Fin Knowledge Management

Control, update, and maintain all of the knowledge sources Fin learns from in one centralized location with the Knowledge Hub. That means Fin will always have access to the latest information as your business changes and grows.

A screenshot of the UI showing the Knowledge sources feature

Knowledge sources

Fin instantly learns from a variety of public and private knowledge sources, including Help Center articles, internal support content, PDFs, and URLs.

2

Behavior

Instantly implements the correct behaviors and policies

Fin speaks in your tone of voice, is fluent in 45 languages, and follows your guidelines, policies, and procedures to deliver the best customer experience every time. You can always rely on Fin to make the right decision.

/New, available now
A screenshot of the UI showing the Tone of voice feature

Tone of voice

Fin can incorporate your TOV preferences, allowing you to customize your own or choose from a selection like professional, friendly, humorous, and more. You can also choose how long Fin’s answers are, from shorter and concise, to longer and more thorough.

A screenshot of the UI showing the Multilingual feature

Multilingual

Fin will automatically detect and resolve issues in more than 45 languages, giving you complete control over which of those languages you would like to support.

/Coming soon in q4
A screenshot of the UI showing the Real-time translation feature

Real-time translation

Fin can translate support content written in any language and use it to generate replies that match the customer's language—no localization required.

Illustration
/Coming soon in q4
A screenshot of the UI showing the Fin Guidance feature

Fin Guidance

Fin can learn your support policies or procedures in natural language, and follow the correct processes with every customer.

/New, available now
A screenshot of the UI showing the AI Category Detection feature

AI Category Detection

Fin automatically detects questions about topics you choose, such as refunds or cancellations, and will follow Workflows or route them to the appropriate team.

A screenshot of the UI showing the Workflows for Fin feature

Workflows for Fin

Add Fin to Workflow automations to triage, route, and generate answers when customers take specific actions—directly from Workflows no-code visual builder.

A screenshot of the UI showing the Custom Answers feature

Custom Answers

Fin can use pre-configured answers with embedded rich media like maps, external data, and custom actions to give you complete control over exactly how specific, high priority or critical questions should be answered.

A screenshot of the UI showing the Targeting & scheduling feature

Targeting & scheduling

Fin can show up for your customers how and when you decide based on audience, region, channel, and more—helping you control and maintain your support availability even when you're out of office.

A screenshot of the UI showing the Human handoff feature

Human handoff

Configure how and when Fin triages conversations or hands off to your human support team. Fin will always automatically handoff when that is the safest option for the customer.

3

Actions

Personalizes the experience for every customer

Fin can access information from your data sources and systems in order to personalize its service for your customers, doing everything it can to efficiently solve problems across your support channels.

/New, available now
A screenshot of the UI showing the Data and actions feature

Data and actions

Connect Fin to your data sources and systems so it can retrieve and update information, enabling it to provide more personalized answers and perform complex tasks on behalf of your customers.

/New, available now
A screenshot of the UI showing the Personalized answers feature

Personalized answers

Fin can retrieve customer information from external systems in order to personalize each customer experience, providing information related to customer accounts, purchases, subscriptions, and more.

/Coming soon in q4
A screenshot of the UI showing the Tasks feature

Tasks

Fin can update information on external systems in order to resolve more complex and time-consuming requests, helping your customers update billing information, cancel orders, change reservations, and more.

/New, available now
A screenshot of the UI showing the Action templates feature

Action templates

Connect Fin to the data sources you use every day—no code required—with templates for Stripe, Shopify, Statuspage, and more. AI-generated templates help you identify the most valuable actions to set up first.

Illustration
A screenshot of the UI showing the Content targeting feature

Content targeting

Fin can deliver content to customers based on their plan, location, brand, and more. You can specify what content is relevant for specific groups of customers, to ensure they only receive relevant answers.

/New, available now
A screenshot of the UI showing the Fin over email feature

Fin over email

Fin delivers instant support via email, making sure to filter out phishing attempts, spam, and other threats. Fin uses all of the email conversation history, and structures every answer specifically for the email channel.

A screenshot of the UI showing the Fin over WhatsApp & SMS feature

Fin over WhatsApp & SMS

Fin delivers conversational support across WhatsApp and SMS, to keep the customer experience consistent across all your channels.

A screenshot of the UI showing the Fin over social feature

Fin over social

Deliver automated conversational support experiences consistently on your social channels, including Facebook and Instagram.

/Coming soon in q4
A screenshot of the UI showing the Fin-first Messenger feature

Fin-first Messenger

The new Messenger integrates Fin 2, Workflows, and your team to help you deliver high quality, conversational support experiences instantly.

4

Insights

Provides the data you need to improve quality and performance

AI-generated insights give you the tools to monitor and improve quality and performance across your entire support organization.

/New, available now
A screenshot of the UI showing the Holistic reporting feature

Holistic reporting

Get visibility into the overall health of your entire support organization with a unified view of human and AI support in one detailed report.

/Coming soon in q4
A screenshot of the UI showing the AI Generated CSAT feature

AI Generated CSAT

For the first time ever, you can get CSAT scores across 100% of your customer conversations using AI-generated analysis.

/Coming soon in q4
A screenshot of the UI showing the Conversation quality feature

Conversation quality

Quickly identify areas of poor quality across your entire support operation to determine areas where you need to improve content or train your agents.

A screenshot of the UI showing the Fin overview report feature

Fin overview report

Analyze how Fin is impacting your resolution rates and CSAT, with a complete overview of Fin’s performance in a pre-built, customizable report.

A screenshot of the UI showing the Fin custom reporting feature

Fin custom reporting

Build your own Fin performance and quality report with new chart styles, drag and drop chart building and chart drill-in to dive deeper into performance.

A screenshot of the UI showing the Fin unresolved questions report feature

Fin unresolved questions report

Get actionable insights on unresolved questions and pinpoint areas for improvement in your support content or AI agent setup.

A screenshot of the UI showing the Conversation monitoring feature

Conversation monitoring

Easily review Fin conversations in real time right within the Inbox.

Performance

Performance beyond what's humanly possible

Customer queries resolved by Fin since March 2023:
8,259,849

Fin was the first AI agent in customer service, and its record speaks for itself.

Fin resolution rate distribution

Intercom customers using Fin see unparalleled resolution rate performance.

Fin Accuracy Rate
99.9%

Accuracy rate based on internal estimate from manually sampled real-world conversations.

Fin Average Resolution Rate over time
Intercom’s AI group runs hundreds of experiments to consistently improve Fin’s resolution rate—meaning you will always get the very best performance.

Customer Stories

“The results we have seen with Fin are groundbreaking, double-digit gains in engagement and resolution rates. Never have I seen a piece of technology so seamless to integrate. Just a few clicks and you suddenly have a new 24/7 teammate.”

73%

Reduction in close time

42%

AI Resolution Rate

1 / 3
Sam Forde
Sam FordeMerchant Support Manager at Zapiet

AI Engine
Powered by the
Intercom AI Engine
The patented Intercom AI Engine™ allows Fin to refine every query, optimize every response and validate the quality of each answer. That's why Fin is the only AI agent that can balance industry-high resolutions with industry-low hallucinations. That makes Fin the highest performing and safest AI agent in customer service.
/ Phase 1
Refine the query

In order to optimize the accuracy of an answer that an LLM generates, the inputs the LLM receives must be refined for comprehension. The clearer and more understandable the query, the better the output.

Specification

  • 1.1

    Check for safety and relevance

  • 1.2

    Optimize query comprehension

  • 1.3

    Check for Workflows automation

  • 1.4

    Check for Custom Answer

/ Phase 2
Generate a response

Once a query has been checked and optimized the next stage is to generate a response using the LLM. For this task, the Intercom AI Engine™ has been designed to use a bespoke and enhanced retrieval augmented generation architecture (RAG for short).

Specification

  • 2.1

    Optimize retrieval

  • 2.2

    Integrate and augment

  • 2.3

    Generate response

/ Phase 3
Validate accuracy

In the final step of the process, the Intercom AI Engine™ performs checks to understand whether the output from the LLM meets the necessary response accuracy and safety standards.

Specification

  • 3.1

    Validate the response

  • 3.2

    Respond to customer

/ always-on
Engine optimization

To calibrate and enhance engine performance, the Intercom AI Engine™ has advanced integrated tools that help optimize answer generation, efficiency, precision, and coverage.

Specification

  • 4.1

    Fin customization and control

  • 4.2

    AI analytics and reporting

  • 4.3

    AI recommendations

/ safeguarding
AI safety and security

Intercom has implemented state-of-the-art security measures to protect Fin against a wide range of LLM threats, including those identified by the OWASP LLM Top 10. By consistently testing a variety of high-end LLMs, and deploying rigorous internal controls, security protocols, and safeguards, Fin is able to achieve the highest level of security and reliability while avoiding potential limitations and threats.

Specification

  • 5.1

    Fin AI Security

  • 5.2

    Regional hosting

  • 5.3

    Compliance

    AICPA SOC 2 Type II
    ISO 27001, ISO27701, ISO 27018 and ISO42001
    ISO 27018
    ISO 27701
    GDPR
    CCPA
    HIPAA
  • 5.4

    Third-party LLM data usage, transfer and storage

Generations
A generation ahead of the competition
The effectiveness of automated frontline support has always been limited: from the canned answers of first generation manual custom bots, to the unpredictability of second generation LLM-wrapper-based AI agents.

Fin 2 is a new generation of AI agent because it is the first to deliver on the promise of human-quality automated support. That’s why Fin 2 represents a step change for the customer service industry.

Gen 1
2018 - 2023
Gen Image
First there were chatbots
System
Rules-based system
Performance
Low resolution rate
Capabilities
Uses pre-configured responses
Gen 2
2023 - 2024
Gen Image
Then generative AI
System
LLM-wrapper based bot
Performance
Improved resolution rate with high hallucinations
Capabilities
Answers simple knowledge-based questions Limited to single source answers
Gen 3
2024
Gen Image
Now there’s Fin 2
System
The patented Intercom AI Engine™
Performance
Industry high resolution rate and low hallucinations
Capabilities
Answers more types of complex questions across channels Uses multi-source generative answers to construct more thorough responses Adapts to support the team's tone of voice Implements support policies when needed Personalizes answers with contextual customer information Takes action in external systems

Customer Stories
With Fin, we've achieved a 50% resolution rate on customer queries that used to require manual support, like ‘What's the due date of my invoice?’ Our customers are now finding answers on their own, it's brilliant.”
Jens NäsholmHead of Automated Support and Tooling at Tibber

Features
More capabilities than any other AI agent
An ink based illustration depicting the build of the Fin AI Engine
Fin 2
Zendesk Bot
Salesforce Einstein

Knowledge

Behavior

Actions

Insights

Pricing
$0.99 per resolution for human-quality service
Fin 2 is the most cost-effective way to scale your support. You only pay for the outcome you care about most—resolved conversations.