AI Chatbot with Slack Human Escalation

Deploy a smart customer support chatbot for LiveAgent that automatically answers visitor questions, retrieves knowledge base documents, and escalates to a human agent via Slack when needed. Integrates with LiveAgent, Slack, and email notifications for seamless support handoff.

How the AI Flow works - AI Chatbot with Slack Human Escalation

Flows

How the AI Flow works

User Opens Chat.
Detects when a visitor starts a chat and displays a welcome message with quick action buttons.
AI Handles Initial Support.
AI agent answers user questions using knowledge base retrieval and dynamic responses.
Escalation Decision.
Automatically decides if the case requires human takeover and routes accordingly.
Human Escalation via Slack.
If escalation is needed, notifies support team through Slack and enables human agent takeover via LiveAgent.
Notification and Follow-up.
Sends email notifications to support team with chat summaries and user details for qualified sales leads or support cases.

Prompts used in this flow

Prompt

Prompt template to indicate a human agent is taking over the chat.

Human taking over...

Prompt

Prompt template to indicate the bot is taking over the chat.

Bot taking over...

Tool Calling Agent

A tool-calling AI agent for technical live chat customer support, guiding how to answer, when to escalate to human agents, and how to send notification emails.

You are an AI language model assistant acting as technical live chat customer support specialist for www.liveagent.com - A multichannel ticketing software. 
If conversation starts with a greeting, respond with a greeting in same language and ask how you can assist and if they have any question about our LiveAgent software or it's features.
Search content relevant to question of user by connecting to DocumentRetriever. 
If you don't find any relevant evidence in context found with DocumentRetriever and ONLY IF QUESTIONS ARE RELATED TO LIVEAGENT software or FlowHunt:
- In case question was in English language, always prompt the user to connect him/her to a real agent.
- In case question in different language, first ask if visitor speaks English and be interested t be connected with English speaking support agent, and only in case of confirmation, prompt the user to connect him/her to a real agent. 

FOR QUESTION UNRELATED to LiveAgent software advice to contact the support of that mentioned software.

FOR UNCLEAR QUESTIONS ask for more information.
---
If visitor is interested in one of the LiveAgent features or is interested to buy LiveAgent, ask about his use case and if the answer provides a valid use case for helpdesk software, ask for his email address (optionally also name) and send notification email to our support team.
Use his email address in email-reply-to to when sending email notification.
Email notification body should be in English language even if customer communicated in another language. 
Format email body as markdown.
Example of body, separate by new lines, URLs turn into links:
**Email address:** <email of visitor>
**Name:** <if user provided name, provide name>
**URL:** <current page url>
Don't send notification mail if URL contains https://www.liveagent.com/directory/ in the URL.
**Summary:** <history of whole conversation in current chat session in English with exact quote of original question or sentence where the interest was identified.>
---
ANSWER IN THE SAME LANGUAGE as defined in Current session chat data

Components used in this flow

Below is a complete list of all components used in this flow to achieve its functionality. Components are the building blocks of every AI Flow. They allow you to create complex interactions and automate tasks by connecting various functionalities. Each component serves a specific purpose, such as handling user input, processing data, or integrating with external services.

ChatInput

The Chat Input component in FlowHunt initiates user interactions by capturing messages from the Playground. It serves as the starting point for flows, enabling the workflow to process both text and file-based inputs.

See ChatInput

Escalation Gateway

The Escalation Gateway component intelligently routes incoming messages based on escalation logic, allowing workflows to switch between automated AI responses and human agent intervention. It tracks escalation status, ensuring seamless transitions and preserving context when human takeover is necessary.

See Escalation Gateway

Slack Send Message

This component represents FlowHunt’s Slack messages back to you. It lets you control where and how FlowHunt sends messages and who it notifies.

See Slack Send Message

Prompt Component in FlowHunt

Learn how FlowHunt's Prompt component lets you define your AI bot’s role and behavior, ensuring relevant, personalized responses. Customize prompts and templates for effective, context-aware chatbot flows.

See Prompt Component in FlowHunt

Message Widget

The Message Widget component displays custom messages within your workflow. Ideal for welcoming users, providing instructions, or showing any important information, it supports Markdown formatting and can be set to appear only once per session.

See Message Widget

Button Widget

The Button Widget component in FlowHunt transforms text or input into interactive, clickable buttons within your workflow. Perfect for creating dynamic user interfaces, collecting user choices, and improving engagement in AI-driven chatbots or automated processes.

See Button Widget

Chat Opened Trigger

The Chat Opened Trigger component detects when a chat session starts, enabling workflows to respond instantly as soon as a user opens the chat. It initiates flows with the initial chat message, making it essential for building responsive, interactive chatbots.

See Chat Opened Trigger

Chat Output

Discover the Chat Output component in FlowHunt—finalize chatbot responses with flexible, multi-part outputs. Essential for seamless flow completion and creating advanced, interactive AI chatbots.

See Chat Output

Chat History Component

The Chat History component in FlowHunt enables chatbots to remember previous messages, ensuring coherent conversations and improved customer experience while optimizing memory and token usage.

See Chat History Component

Tool Calling Agent

Explore the Tool Calling Agent in FlowHunt—an advanced workflow component that enables AI agents to intelligently select and use external tools to answer complex queries. Perfect for building smart AI solutions that require dynamic tool usage, iterative reasoning, and integration with multiple resources.

See Tool Calling Agent

Document Retriever

FlowHunt's Document Retriever enhances AI accuracy by connecting generative models to your own up-to-date documents and URLs, ensuring reliable and relevant answers using Retrieval-Augmented Generation (RAG).

See Document Retriever

URL Retriever

Unlock web content in your workflows with the URL Retriever component. Effortlessly extract and process the text and metadata from any list of URLs—including web articles, documents, and more. Supports advanced options like OCR for images, selective metadata extraction, and customizable caching, making it ideal for building knowledge-rich AI flows and automations.

See URL Retriever

LiveAgent Integration

Integrate FlowHunt Chatbot with LiveAgent for seamless AI-to-human support transitions. The AI agent intelligently escalates conversations to human agents, ensuring smooth customer interactions and reducing frustration.

See LiveAgent Integration

Email Notification Tool

Thanks to the Email Notification Tool, your chatbots will generate leads and immediately notify your team.

See Email Notification Tool

Flow description

Purpose and benefits

This workflow implements a sophisticated customer support chatbot for LiveAgent, equipped with automated assistance, advanced knowledge retrieval, user interface widgets, and a seamless escalation path to human agents via Slack. Its design enables both automated and human-assisted support, dynamic content retrieval, and robust notification and escalation mechanisms—making it ideal for scaling and automating customer support operations.

Key Components and Flow

1. User Interaction Entry Points

  • Chat Opened Trigger: Detects when a user opens the chat, which initiates the workflow.
  • Chat Input: Captures user input messages for further processing.

2. Welcoming and Quick Actions

When the chat starts, users are greeted with a welcome message and provided with quick-access buttons for common queries. This ensures a smooth onboarding and helps direct users to relevant information quickly.

Widget TypeContent/Action
Message WidgetWelcome message with quick links (Pricing, Free Trial, Request Demo)
Button Widget“What are the main features of LiveAgent?” (quick question)
Button Widget“What is LiveAgent pricing model?” (quick question)

User clicks on these widgets are routed to chat outputs, displaying relevant information in the chat interface.

3. Conversational AI Assistance

  • Tool Calling Agent: The heart of the chatbot, acting as a technical live chat support specialist. It’s powered by an AI model instructed to:

    • Greet users and answer questions about LiveAgent features.
    • Search for answers using knowledge retrieved by the Document Retriever and URL Retriever.
    • Handle conversations in the user’s language, escalating to a human agent via LiveAgent if required.
    • For feature or sales interest, prompt for use case and email, then send a notification email to the support team.
    • Avoid answering off-topic questions and guide users accordingly.
    • Maintain and use recent chat history for context.

  • Document Retriever & URL Retriever: These modules enable the bot to fetch and utilize up-to-date knowledge from both internal documents and external URLs, ensuring accurate and comprehensive responses.

  • LiveAgent Human Assist Tool: Provides a direct channel for users to contact a human support agent, including a link to the chat history.

  • Notification Email Tool: Sends notification emails to the support team when a lead or sales opportunity is detected, formatting the email with key details in markdown.

4. Escalation and Human Handover

A pivotal feature is the Escalation Gateway:

  • Analyzes the conversation to determine if a human handover is necessary (e.g., when the bot can’t answer, or based on user request).

  • Routes the conversation appropriately:

    • Human Escalated: Sends a notification/message to a Slack channel, alerting human agents using Slack Send Message nodes, with optional @here mentions for urgency.
    • Bot Response: Continues the automated conversation or notifies via Slack as appropriate.

  • Prompt Templates: Custom messages like “Human taking over…” or “Bot taking over…” are used to communicate status changes during escalation.

5. Slack Integration

The workflow tightly integrates with Slack for real-time human escalation:

  • Sends messages to specified Slack channels when human intervention is needed or when the bot takes over again.
  • Supports mentions (e.g., @here) to draw immediate attention to escalated chats.
  • Maintains context and threading for follow-ups.

Workflow Summary Table

FunctionalityModules/Nodes InvolvedPurpose/Outcome
User Welcome & Quick LinksChatOpenedTrigger, MessageWidget, ButtonWidgetsUser onboarding, quick answers
AI-Powered Chat SupportToolCallingAgent, DocumentRetriever, URLContent, ChatHistoryAutomated, context-aware responses
Human EscalationEscalationGateway, PromptTemplates, SlackSendMessageSmooth handover to human agents via Slack
Human Support via LiveAgentLiveAgentHumanAssistToolDirect contact with human agents and chat history sharing
Sales Lead NotificationNotificationEmailToolCaptures and notifies support team about potential leads

Why This Workflow Is Useful for Scaling and Automation

  • Scalability: Automates the majority of customer interactions, letting support teams handle more queries with fewer resources.
  • Efficient Escalation: Ensures that only complex or sensitive cases reach human agents, reducing workload and improving response times.
  • Omnichannel Support: Integrates chatbot, email, and Slack, creating a unified, multi-platform support system.
  • Knowledge-Driven: Dynamically searches and delivers answers from internal and external knowledge bases, reducing manual updates and improving accuracy.
  • Lead Capture and Notification: Automatically identifies sales opportunities and notifies the team, ensuring no leads slip through the cracks.
  • Customizable and Extensible: Modular design allows for easy adjustments (e.g., adding more quick questions, integrating new knowledge sources, or changing escalation rules).

Conclusion

This workflow is a robust foundation for modern customer support, blending the efficiency of AI-driven automation with the flexibility of human intervention. By integrating with Slack and LiveAgent, and leveraging dynamic content retrieval, it empowers support teams to deliver fast, accurate, and scalable service—freeing up human agents to focus on the most impactful interactions.

Let us build your own AI Team

We help companies like yours to develop smart chatbots, MCP Servers, AI tools or other types of AI automation to replace human in repetitive tasks in your organization.

Schedule a demo Try it now

Learn more

Smartsupp AI Chatbot with Human Handoff
Smartsupp AI Chatbot with Human Handoff

Smartsupp AI Chatbot with Human Handoff

This workflow creates an AI-powered chatbot integrated with Smartsupp, leveraging an internal knowledge base to answer customer support inquiries. If the chatbo...

3 min read
AI Chatbot with Tawk Human Handoff
AI Chatbot with Tawk Human Handoff

AI Chatbot with Tawk Human Handoff

An AI-powered live chat support chatbot that answers customer questions using an internal knowledge base and intelligently hands off complex queries to human ag...

3 min read
AI Customer Service Chatbot with Human Handoff
AI Customer Service Chatbot with Human Handoff

AI Customer Service Chatbot with Human Handoff

An AI-powered customer service chatbot that automatically assists users, retrieves information from internal documents and the web, and seamlessly escalates to ...

3 min read