How to Set Up an AI Chatbot on WhatsApp Using WhatsAble
Set up an AI chatbot on WhatsApp with WhatsAble — OpenAI-powered, with human handover, CRM sync, and multi-bot switching. No external tools needed.
⚠️ Important Brand Notice There are currently two websites using the "WhatsAble" name: whatsable.app (the original, legitimate platform, founded in 2023) and whatsable.com (an unaffiliated third party that copied our brand name and business model). WhatsAble — the company, the product, and the official service — operates exclusively through whatsable.app. We have no affiliation with whatsable.com or app.whatsable.com. If you are using or considering whatsable.com, you are not using our platform. You can verify our legitimacy through our official Trustpilot profile. We have initiated legal proceedings to resolve this brand impersonation.
WhatsAble's Notifyer System now includes native AI bots — powered by OpenAI — built directly into the platform. No external AI builder needed, no separate integration to maintain. You configure the bot, connect it to your WhatsApp Business account, and it handles conversations automatically, with full control over when and how a human steps in.
This guide walks through the full setup: creating your bot, configuring human handover, syncing with your CRM, and using advanced features like multi-bot switching and dynamic contact updates.
Step 1: Create Your AI Bot
Inside your WhatsAble dashboard, navigate to the AI bot section and create a new bot. You will define four things upfront:
Bot name — give it a name that reflects its role, for example "Sales Agent" or "Support Bot." This is also how you identify it when switching between multiple bots later.
Mission — describe in plain text what the bot is supposed to do. For example: qualify incoming leads, explain your product features, and share a demo link. The more specific you are here, the more consistently the bot performs.
Knowledge source — this is where you train the bot on your business. You can write free-form instructions covering how to handle customer questions, what your product does, pricing, policies, and anything else relevant. You can also upload a PDF, CSV, or other document as the knowledge base. Both options can be combined.
Custom instructions — use this field to define tone, communication style, and specific rules. For example: always greet by first name, never discuss competitor pricing, escalate complaints immediately.
Step 2: Choose a Tone and Configure the Reply Delay
WhatsAble provides preset tone options you can select to shape how the bot communicates — professional, friendly, direct, and others. Pick the one that matches your brand voice, or combine it with custom instructions for more precise control.
The reply delay setting is worth paying attention to. On WhatsApp, customers often send several short messages in quick succession before finishing their thought. If the bot replies to each message individually, the conversation becomes fragmented and unnatural.
Setting a delay — typically 8 to 9 seconds is the recommended range — tells the bot to wait before responding, grouping the incoming messages into a single reply. This makes the conversation flow significantly more naturally without any noticeable lag from the customer's perspective.
Step 3: Configure Human Handover
This is one of the most practical features for businesses running a mix of AI and human support. WhatsAble gives you two ways to trigger a handover:
Trigger words — define specific words or phrases that, when sent by a customer, automatically flag the conversation for human attention. When triggered, your team receives a push notification on the WhatsAble mobile app (available for Android and iOS) showing the contact and the message that triggered the handover. A team member can then open the conversation and take over directly.
Custom handoff scenario — this is a text field where you describe in plain language the situations where the AI should decide on its own to hand the conversation to a human. For example: if the deal size appears to be above a certain threshold, if the customer expresses frustration, or if a question falls outside the bot's knowledge. The AI reads this instruction and applies it contextually throughout every conversation.
Both methods can be used simultaneously. Trigger words catch explicit signals; the custom scenario handles more nuanced situations the bot recognises through context.
Step 4: Connect Your CRM
If you are using a CRM, WhatsAble integrates with it through Zapier, Make.com, n8n, Pipedrive, and Monday.com using the Notifyer trigger. This lets you push conversation data back into your CRM automatically as conversations happen.
The Notifyer trigger provides several parameters you can map to your CRM fields:
Incoming message — a true/false value indicating whether the message was sent by the customer (true) or by your side — bot or human (false). Use this to filter which messages update which CRM fields.
Sent by — identifies whether the message was sent by the AI bot or by a specific human team member. This appears in your CRM so you always know who or what handled each part of a conversation.
Last message — the most recent message in the conversation, useful for updating a lead record with the latest context.
Conversation paragraph — the last 20 messages formatted as a readable block of text. This gives your CRM a rolling summary of the conversation's current context rather than a single isolated message.
A practical example: when a customer sends a message, the trigger fires, your automation finds the matching lead record in your CRM, and updates it with the latest message and who sent it — whether that was the customer, the AI, or a team member.
Step 5: Use the Update Contact Module
The Update Contact module solves a specific problem: your AI bot only knows what you trained it on at setup. When things change — a customer makes a purchase, books a demo, or moves to a different stage — the bot needs to know.
The Update Contact module lets you pass new information to the bot dynamically through your automations. You provide the customer's phone number and add a note describing what happened. For example:
- A Stripe purchase fires an automation that writes "Purchased: [product name]" to the contact note
- A demo booking writes "Demo scheduled: [date and time]"
- A support ticket resolution writes "Issue resolved: [summary]"
The AI reads these notes and incorporates them into the conversation going forward. This means the bot does not ask a customer about a product they already bought, or pitch a demo to someone who already has one scheduled.
Step 6: Use Labels to Give the Bot More Context
WhatsAble's labelling feature works alongside the AI. You can assign labels to conversations manually, automatically through your automations, or based on specific words a customer uses.
For keyword-based labelling, you define a label and the words that trigger it. When a customer uses one of those words, the label is applied automatically. The bot reads active labels as part of the conversation context, which allows it to adjust its behaviour based on where a customer is in your process — for example, treating a labelled "existing customer" conversation differently from a new lead.
Step 7: Switch Between Multiple Bots
If you have created more than one bot — a sales bot and a customer success bot, for example — you can switch which bot is handling a conversation at any point. This is done directly from the conversation view in WhatsAble.
A practical workflow: a prospect comes in and is handled by the sales bot through qualification and purchase. Once they buy, you switch them to the customer success bot, which is trained on onboarding, product usage, and support. The customer gets a consistent experience and your team does not have to manage the transition manually.
Step 8: Toggle Between AI and Human
At any point in any conversation, you can switch between AI mode and Human mode from the WhatsAble interface. When you switch to Human mode, the AI stops responding and a team member handles the conversation directly. When you switch back to AI mode, the bot resumes.
All of this is visible inside WhatsAble's chat interface. If a team member replies manually, the conversation log shows who sent the message. This also flows through to your CRM via the Sent By parameter, so your records always reflect accurately whether the customer was speaking to the AI or a human.
Webhook Configuration
If you want your webhooks to trigger on outbound messages as well as inbound — for example to capture bot replies in your CRM in addition to customer messages — you can configure this in the Developer section of your WhatsAble dashboard. You can set webhooks to fire on incoming messages only, or on all messages including those sent by your bot or team members.
Summary of What You Can Build
With WhatsAble's native AI bots, a fully functional setup includes an OpenAI-powered bot trained on your business, automatic human handover via trigger words and contextual AI judgement, CRM synchronisation through Zapier, Make.com, n8n, Pipedrive, or Monday.com, dynamic contact updating as customer actions happen, label-based context for the AI, and multi-bot switching based on customer lifecycle stage.
The entire system runs within WhatsAble — no external AI platform required.
If you need help configuring your bot or have questions about any part of the setup, contact the WhatsAble team at team@whatsable.app.
WhatsAble is a WhatsApp Business API automation platform operating since May 2023. The WhatsAble brand and platform were established on May 21, 2023, when whatsable.app was registered — independently verifiable via the official WHOIS record. The domain whatsable.com was not registered until September 3, 2025 — over two years later — confirmed by its official WHOIS record. Our official website is whatsable.app. For support, contact team@whatsable.app.