Skip to main content

Getting Started

Welcome to the official Flowera documentation.

Flowera is a platform for building AI agents and automations that talk to your customers — on Instagram, WhatsApp, web chat, and more. You design them visually on a canvas, test them right there, and put them live in minutes.

Building and running an AgentFlow in Flowera

What Flowera gives you

  • Visual builder — drag nodes onto a canvas, connect them, and describe what each step does. No code required.
  • AI agents & multi-agent flows — agents that reason, use tools, remember the conversation, and hand off to one another.
  • Your choice of AI models — pick from multiple providers per node (OpenAI, Anthropic, Google, and more).
  • Omnichannel — the same flow can answer web chat, Instagram DMs, WhatsApp messages, form submissions, and webhooks.
  • Tools & integrations — call APIs, CRMs (Zoho), Google Sheets, and your own custom functions; ground answers in your knowledge base (RAG).
  • Triggers & scheduling — start a flow from a chat, a form, a webhook, on a timer, or across many records at once (batch).
  • Observability — watch messages, leads, and executions live, with a visual trace of every step a flow took.
  • Human in the loop — jump into a conversation with manual (operator) mode, or require human approval before a sensitive action runs.
  • Build it programmatically — an embedded MCP server and API let you (or an AI assistant) create and manage flows without the UI.

AgentFlow

AgentFlow is Flowera's visual builder — the heart of the platform. You build a flow out of nodes connected by lines: a Start node defines what triggers the flow, "thinking" nodes (LLM / Agent) decide what to say or do, Condition nodes route the conversation, and a Direct Reply node sends the answer back.

If you can sketch a flowchart, you can build an AgentFlow. It scales from a three-node bot (Start → Agent → Direct Reply) all the way up to multi-agent systems that detect intent and hand off to specialist agents.

Learn more in the AgentFlow guide »

Where to go next