ChatHub – Omnichannel chat management & drag-and-drop bots with AI

Workflows are ChatHub’s core strength. Connect message, menu, question, condition, product sale, top-up, API, AI and agent handoff steps on a visual canvas. Branches describe what happens after a customer choice or a failed step.

Describe the starting situation, the task and the desired outcome to AI. It reads the workflow draft and conversation to help propose and create suitable steps. Continue the discussion, review the diagram and edit the steps before saving. Configure an AI provider first.

The template library covers digital product sales, COD orders, after-sales care, top-ups and appointments. Creating from a template builds the steps and leaves the workflow disabled so you can review content and configure services before enabling it.

Customise greetings, menus, branches and shared variables for your business. Extend the template with information collection, external API calls or a handoff to your team.

The automated digital shop template creates a store inside chat. Its menu includes purchases, top-ups, balance, orders, guidance and support. Customers choose a digital product, confirm their purchase and enter a quantity; the bot charges the wallet, delivers according to the product settings and returns them to the menu.

Insufficient balance leads to bank top-up through Pay2S or Binance Pay. A top-up only credits the wallet: customers choose and confirm a new purchase themselves. Out-of-stock branches offer another product, while other failures lead to support. Customers can also look up recent orders, check their status and retrieve delivered content.

Prepare digital stock, enable automatic product delivery and configure the template’s Pay2S/Binance Pay integrations. In the Variable tab, customise shop name, support contact and hours, purchase notes, quantity limits and order history limits. Test the flow, then enable the workflow and its new-customer greeting; disable that greeting on any competing workflow.

Self-service actions return to the menu. The bot can restart when a customer returns after the session expires and yields when an agent takes over, combining automated sales with human assistance.

ChatHub supports Facebook Pages, personal Zalo, Telegram Bot, personal Telegram and the website Chat Plugin. Messages from connected accounts appear in one system so agents can work together while keeping track of each conversation’s channel.

Connect Facebook Pages to bring messages sent to your page into the shared inbox. Your team can follow and reply to Page conversations alongside other channels in ChatHub, making it easier to serve customers who contact your business in different places.

Inside a conversation, agents can add tags and internal notes, assign an agent or team, and update customer profiles. Agents can take over from the bot for individual support. The website Chat Plugin also supports images, files and a pre-chat contact form.

Embed the Chat Plugin script on your website to add a support button. Visitors can open the chat and ask questions without leaving the page, sending text, images or files. Customise colours, the title, greeting and launcher position to fit your brand.

Messages arrive in the shared ChatHub inbox for agents to receive, assign and answer. Replies appear in the widget without a page reload, and history is retained for the visitor session so customers can continue when they return. Optional settings include a pre-chat contact form and multiple conversations.

Connect live chat with configured workflows to welcome customers, offer a consultation menu or hand the conversation to an agent. Your website becomes another entry point into the same customer service and sales process.

Add a “Reply with AI Agent” step to a workflow so the bot can compose and send replies to customers. AI reads your instructions and recent messages to continue the conversation in context. Configure the step to wait for an answer to support multi-turn conversations.

Choose which knowledge sources AI can use: product catalogues, prices, saved replies, help documents or orders belonging to the customer in the conversation. For example, product questions can be answered using your configured catalogue and documentation. Response quality depends on the supplied data and selected model.

When a conversation needs a specific action, AI Agent can select an enabled workflow from your permitted list, such as a purchase or top-up flow. Combine human handoff steps and error branches in the workflow so your team can handle cases requiring direct assistance.

Before enabling it, configure the AI provider, model and API key in Settings, prepare knowledge sources and instructions, and test conversation scenarios. AI usage is paid directly to your provider. Inbox AI suggestions remain drafts for an agent to review and send.

Agents can request a draft based on conversation history and configured knowledge sources such as prices, products and documents. The suggestion appears in the composer for review before sending. Rewriting helps polish an existing draft, while message translation helps agents communicate across languages.

An AI Agent step within a workflow can respond using selected instructions and data sources. Settings cover AI providers, knowledge documents and website content collection. Results depend on the model, knowledge and configuration; an AI service must be configured before use.

Appointments include day, week and month views, agent filters and pending or confirmed states. Working hours and exceptions let the system check slot availability when making a booking.

Appointment steps in workflows bring booking into the conversation. This helps service teams turn an enquiry into a specific appointment and track their schedules in the same system.

Campaigns let you select customers by tags and conditions, personalise messages with variables, preview content, send a test and schedule delivery. Teams can prepare relevant follow-ups for a chosen audience.

Reports cover conversations, messages, contacts, appointments, response speed, agent activity, automation and AI usage. Combined with roles, team management and activity logs, these views help operators track workload and refine their service.