Fin is an AI-powered Customer Agent built for support, sales, and CX teams that want to automate customer conversations across chat, email, phone, and messaging channels without hiring additional headcount. It is aimed at mid-market and enterprise businesses running customer service and inbound sales operations, particularly companies already using helpdesks like Intercom, Salesforce, or HubSpot. The core workflow revolves around connecting Fin to a company's knowledge base, policies, and backend systems, then letting it autonomously resolve tickets, route conversations, or hand off to a human agent when needed. Fin positions itself as a single agent that spans the full customer journey rather than a narrow chatbot limited to FAQ deflection.
What it does
Fin is a conversational AI agent designed to handle customer service and inbound sales interactions on behalf of a business. It connects to a company's existing systems and knowledge sources, then answers customer questions, takes actions like updating records, and escalates to human agents when a conversation requires it. Fin works across web chat, email, SMS, WhatsApp, and phone, and can be deployed alongside an existing helpdesk or through Intercom's own inbox and ticketing system. The product is built for CX leaders, support operations teams, and RevOps or sales teams who need an AI agent that can be configured on tone, behavior, and business rules without engineering involvement.
Key capabilities
- Cross-channel deployment: Fin operates across live chat, email, SMS, WhatsApp, and voice, giving businesses one agent for multiple support surfaces instead of channel-specific bots.
- No-code configuration: Tone, behavior, and knowledge sources can be set up by non-technical team members through procedures, data connectors, guidance, and categories rather than custom code.
- Action-taking and system integration: Fin connects to external systems via API and data connectors to look up records and update information, not just answer questions from static content.
- Auto-triage and routing: Conversations that fall outside Fin's scope are automatically routed and handed off directly into the human agent's existing inbox.
- Add-on performance tools: Optional Operator and Copilot add-ons provide AI-driven performance monitoring (CX Score, AI Topics, Trends, Incident Detection) and an in-inbox AI assistant for human agents.
- Helpdesk flexibility: Fin can run on top of an existing helpdesk (Salesforce, HubSpot, etc.) or be paired with Intercom's native inbox, ticketing, and public help center.
Pricing
Fin uses outcome-based pricing rather than flat seat licensing. When used with an existing third-party helpdesk, Fin is billed at $0.99 per resolved outcome with a 50-outcome monthly minimum. When paired with the Intercom helpdesk, pricing starts from $0.99 per outcome plus $29 per helpdesk seat per month, with a free trial available. Two optional add-ons are listed separately: Operator (advanced AI analytics and quality monitoring) starts from $99 per month, and Copilot (an AI assistant for human agents) is priced at $35 per user per month. There is no permanent free tier, but a 14-day free trial is offered, and an Early Stage Program provides startups a discount on Intercom plus a year of Fin. Pricing page: View pricing
Editorial review
Fin's outcome-based pricing model is a notable point of differentiation from typical per-seat SaaS pricing, aligning cost with actual resolved conversations rather than agent count — a structure that will appeal to CX leaders trying to tie AI spend directly to measurable results, but one that also requires careful ROI modeling since costs scale with support volume. The claimed customer base (over 12,000 brands) and prominent customers referenced on the marketing pages, including Anthropic, suggest meaningful enterprise adoption, though independent verification of resolution-rate metrics and specific performance benchmarks against competitors should be treated as vendor-reported figures rather than confirmed third-party data. The product's willingness to run on top of third-party helpdesks (Salesforce, HubSpot) rather than forcing a full platform switch is a practical advantage for companies with existing CX infrastructure, while the deeper Intercom-native option adds ticketing, a help center, and outbound tooling for teams open to consolidating their stack. The optional Operator and Copilot add-ons indicate a broader push toward AI-assisted quality management and human-agent augmentation, not just pure automation, which may appeal to teams that want to keep humans in the loop rather than fully replace them. Best-fit users appear to be mid-market and enterprise support and sales operations with sufficient conversation volume to justify per-outcome pricing and the setup effort of connecting knowledge sources and data connectors; smaller teams with low ticket volume or tight per-unit cost sensitivity may find the pricing structure and setup requirements less proportional to their needs. Missing from the public materials are granular details on model accuracy across industries, SLA guarantees, and how "outcome" is precisely defined and audited, all of which prospective buyers should clarify directly during a sales conversation before committing.
