Getting Started
How does the 24-hour demo work? +
After filling out the demo request form, you get instant access to a fully working sandbox configured around your restaurant. You can test the AI agent, generate sample social posts, and explore the dashboard for 24 hours before deciding to activate a paid plan.
Do I need a credit card to start the demo? +
No. The 24-hour sandbox is completely free, no credit card required. You only pay if you choose to activate a plan afterward.
How long does setup take? +
Most restaurants are live within 10 minutes — connecting your WhatsApp Business number and configuring your product catalog and policies is the bulk of the process.
Payments & Billing
What payment methods do you accept? +
We accept Credit Card, OMT Transfer, Bank Transfer, and Cash on Delivery / Bill Me. For OMT and Bank Transfer, you'll enter your transaction reference and upload a receipt for our team to verify, usually within a few hours.
Can I cancel anytime? +
Yes, all plans are month-to-month with no long-term contract. You can cancel at any time from your dashboard or by contacting support.
Broadcast Service
How much does a broadcast cost? +
Your first WhatsApp broadcast is $29 flat. Every additional broadcast after that gets a 10% discount, applied automatically.
Can I use my own contact list? +
Yes — you can submit your own list (paste contacts or upload a file), or choose from our segmented house lists. All lists go through a brief admin review before sending.
What can I send in a broadcast? +
Broadcasts must be strictly commercial — promotions, product updates, event invites, partnership messages, etc. Political content, advocacy messaging, or anything infringing on third-party rights is not permitted. See our
Acceptable Use Policy.
Who approves my broadcast order? +
Our team reviews every order's message and list before anything is sent or charged. Once approved, you'll be invited to complete payment, and we handle the actual sending.
Technical
Does this work with Shopify or WooCommerce? +
Yes, both are supported as connectable integrations. You can also set up your catalog manually if you don't use either platform.
Does the AI support Arabic? +
Yes — the AI automatically detects whether a customer is messaging in Arabic or English and responds accordingly, no configuration required.
Agentic AI — The Technology Behind TableBot
What makes agentic AI different from a regular WhatsApp bot? +
Traditional WhatsApp bots follow rigid "if-this-then-that" paths. If a user deviates from the script, the system breaks. Agentic AI is given a high-level objective — for example, "successfully onboard this new partner" or "verify this transaction invoice" — and dynamically manages the entire multi-turn conversation to reach it. It figures out what questions to ask, processes responses in any order the user provides them, and keeps the interaction moving toward the goal. No script. No breaking.
Can the AI actually take actions, or does it just reply to messages? +
It acts. If a user sends a query or document via WhatsApp, the agent can autonomously decide to trigger a specific API webhook to check an internal database, call a multi-modal sub-agent to parse data from an attached PDF or image, or update a backend CRM or ledger entry instantly based on the conversation text. The user simply texts naturally, while the agent orchestrates the underlying enterprise software stack in the background — invisibly.
What happens if a customer doesn't reply for hours? Does the AI lose context? +
No. WhatsApp is inherently asynchronous — users might reply immediately, or three hours later. Agentic AI maintains a persistent state and a deep memory of the conversation's context. When a user resumes hours later, the agent doesn't restart the script; it evaluates where the process left off, cross-references any backend system changes that occurred in the interim, and adapts its next step accordingly. The conversation picks up exactly where it paused.
What is a Self-Correcting Verification Loop? +
In data-gathering workflows — such as collecting registration details or operational metrics — human inputs are often messy or incomplete. An agentic framework acts as a real-time validator. If a user uploads a blurry photo of a document or provides an invalid data field, the agent recognizes the exception, analyzes what is missing, and immediately asks a precise clarifying question on the chat to resolve the issue on the spot. No manual back-and-forth needed from your team.
Is there really only one AI running everything, or is it more complex than that? +
To the user, they are chatting with a single WhatsApp number. In reality, an agentic architecture allows a front-facing "Interface Agent" to triage the incoming message and route it to specialized internal agents — a Research Agent, an Analytics Agent, a Compliance Agent, and so on. These background agents collaborate, compile the necessary intelligence, and feed the final synthesized response back to the user seamlessly. One number. One conversation. Multiple experts working in parallel behind it.