The visual workflow platform for WhatsApp

Build WhatsApp workflows that do real work.

Design, simulate, version, and deploy bookings, payments, support, and AI-assisted journeys to your own or your clients’ WhatsApp numbers—without building the messaging backend.

No card required · No Meta setup to build or test

Looping animated preview of Fluenta's flow editor: a WhatsApp booking flow keeps growing while the canvas glides sideways. A trigger connects to a button menu, then booking and an AI knowledge agent, then payment, human handoff and contact-tagging nodes join, and a fresh flow starts assembling as the loop continues. A connected phone plays the resulting conversation, including a deposit payment and booking confirmation.

Built on Meta’s official WhatsApp Business Platform

Client-owned numbers and accounts

Versioned publishing

JavaScript, Python, HTTP, and webhooks

Inspect the product

Open a real workflow before you create an account.

Built to be inspected before you trust it. Follow a booking, support, commerce, or qualification journey from the incoming WhatsApp message to the final operational result.

app.fluenta.dev/bots/nova
Editor view of the Fluenta workspace

Drag any of 24 node types onto the canvas, or describe the workflow and let AI draft a starting point.

Live on WhatsApp

Feels personal. Runs automatically.

Every reply your customers see comes from a flow you built: buttons, booking slots and knowledge answers, delivered inside WhatsApp's own chat chrome.

  • Buttons, lists & rich replies

    Native WhatsApp UI, not walls of text.

  • Answers from your own docs

    The bot answers using your approved menus, policies, and FAQs.

  • Booking built in

    Live availability, booking confirmation, calendar feed.

  • Human handoff

    Escalates to your team inbox with full context.

CA

Café Andino

automated repliestyping…

An actual Fluenta template. Install it and swap the copy.

A customer messages Café Andino on WhatsApp asking for a table for four tonight. The Fluenta bot greets her by name and offers three time slots as tappable buttons. She taps 8:30 PM, and the bot confirms the reservation with a summary card carrying the booking reference, then signs off warmly. The conversation loops continuously, with each finished chat scrolling up as history.

More than a chatbot canvas

Build the workflow. Prove it. Then put it live.

Most WhatsApp projects fail between the diagram and production. Fluenta keeps the customer journey, runtime logic, testing, deployment, and operations in one inspectable workflow.

01

Visual by default

Map messages, conditions, bookings, payments, knowledge, and handoff on the canvas.

02

Test before Meta

Run every branch, inspect variables, and trace execution without connecting a number.

03

Code when required

Call an API, run JavaScript or Python, map outputs, and keep explicit error paths.

04

Publish deliberately

Release a validated, immutable workflow version without changing the draft under construction.

Workflow lifecycle

From brief to live operation

01

Design

Start from a template, describe the workflow to AI, or build it node by node.

02

Simulate

Test customer inputs, conditions, variables, media, fallbacks, and handoff paths.

03

Deploy

Connect the business-owned WhatsApp number and publish a validated version.

04

Improve

Measure completion, conversion, handoff, and drop-off against the actual workflow.

Operational outcomes

One workflow can complete the job, not just answer the message.

Bookings and payments

Offer live availability, collect the customer’s details, request a deposit, and create the confirmed appointment.

Explore bookings

Customer support

Answer from approved information, route uncertain questions to a person, and preserve the complete conversation context.

Explore support

Lead qualification

Ask the right questions, score or route the lead, sync the result, and bring in a salesperson only when needed.

Explore qualification

Client delivery

Create isolated workspaces, reuse approved templates, manage access, and deploy to client-owned WhatsApp accounts.

Explore agencies

Built for repeatable client delivery

Stop rebuilding the WhatsApp backend for every client.

Give each client an isolated workspace, start from a reusable workflow, test the complete journey, and deploy to the WhatsApp account the client owns.

  • Separate client workspaces
  • Reusable and versioned templates
  • Owner, editor, agent, and viewer roles
  • Client-owned Meta account and number
  • Portable workflow definitions

Product capabilities

Everything required to run the workflow.

Workflow builder

Messages, questions, conditions, media, lists, bookings, payments, knowledge, handoff, APIs, and code in one graph.

Simulator and debugger

Exercise every branch and inspect variables, call stack, trace, and executed edges.

Versioned publishing

Validate the draft and publish an immutable workflow definition.

AI and approved knowledge

Draft workflows from a prompt and answer customer questions from controlled business sources.

Human operations

Escalate a conversation, assign the ticket, reply with context, and return the customer to the workflow.

Extensibility

Use secrets, signed webhooks, scoped tokens, HTTP requests, JavaScript, Python, and portable JSON.

Ownership and trust

Your channel stays yours.

Each business connects its own WhatsApp Business account and phone number. Meta charges remain on that account, and the workflow can be exported without including its credentials.

Review security
  • Official WhatsApp Business Platform APIs
  • Encrypted secrets at rest
  • Signed inbound and outbound webhooks
  • Workspace roles and tenant isolation
  • Ninety-day default conversation retention
  • Exportable workflow definitions

Start from a working pattern

Open a workflow built for the job.

Explore 9 workflows with their outcome, key nodes, required integrations, and setup detail before you make one your own.

Explore templates

Pricing

Free to prove. Priced when the workflow becomes an operation.

Free

Build and validate a pilot.

$0/mo
  • 1 live WhatsApp number
  • 100 active contacts/month
  • 1 member
  • 3 draft workflows
  • Visual builder, simulator, and debugger
  • Templates and AI workflow drafting
  • Knowledge-grounded answers
  • Versioned publishing
  • Documentation and community support
  • $1 of AI and compute usage included
Start free

Launch

Run one serious WhatsApp operation.

$19/mo
  • 1 live WhatsApp number
  • 1,000 active contacts/month
  • 3 members
  • Unlimited draft workflows
  • Runtime AI and knowledge answers
  • Bookings, payments, and human handoff
  • HTTP and code nodes
  • Campaigns and analytics
  • Email support
  • $10 of AI and compute usage included
Launch a workflow
Most popular

Growth

Scale customer operations.

$49/mo
  • 3 live WhatsApp numbers
  • 5,000 active contacts/month
  • 5 members
  • Unlimited draft workflows
  • Shared operations inbox with assignment
  • Advanced campaigns and funnel analytics
  • Multiple knowledge collections
  • Frontier model tier for runtime AI
  • Priority support
  • $20 of AI and compute usage included
Choose Growth

Compare every plan

FAQ

Answers before you connect.

Can I build before I have a WhatsApp number?

Yes. The canvas, simulator, debugger, templates, and workflow publishing can be used before connecting Meta. A number is required only for live WhatsApp traffic.

Does Fluenta provide a WhatsApp number?

No. Each business connects a number and WhatsApp Business account it owns.

Is Fluenta only no-code?

No. The workflow is visual by default, with HTTP requests, JavaScript, Python, signed webhooks, API tokens, and JSON import/export when deeper integration is required.

Does Fluenta add a markup to Meta’s message fees?

After the pricing model is changed: no. Meta bills its fees directly to the connected business account. Fluenta charges separately for the platform and metered AI or compute usage.

Can an agency manage several clients?

Yes. Agency plans provide isolated client workspaces and defined live-deployment capacity. Each client can retain ownership of its Meta assets.

What happens when AI cannot answer?

Knowledge workflows should follow an explicit fallback path such as clarification, a deterministic response, or human handoff. Fluenta should not invent an answer outside the configured sources.

Build the workflow before you build the backend.

Start with a template or an empty canvas. Test the complete journey without a card or Meta setup.