What is Conduet?

Conduet is a full-stack AI agent builder platform. Design, build, test, and deploy conversational AI agents that can handle customer support, answer questions from your knowledge base, execute multi-step workflows, and integrate with external services — all from an intuitive visual interface.

Who is it for?

Conduet is built for teams that want to create intelligent conversational agents without extensive AI/ML expertise. Whether you are a product team building customer-facing support bots, a developer integrating AI into your application via API, or a business automating internal workflows, Conduet provides the tools you need.

The platform supports two development approaches: a visual conversation flow builder for structured, step-by-step dialogs, and an agentic framework for autonomous agents that reason, use tools, and make decisions on their own.

Key Capabilities

Visual Workflow Builder

Design conversation flows with a drag-and-drop canvas. Connect steps, define branching logic, and build complex multi-turn agents without writing code.

Agentic AI Framework

Build autonomous agents that can use tools, search knowledge bases, make API calls, and handle multi-step reasoning across a conversation.

Knowledge Base with RAG

Upload documents, PDFs, or crawl URLs. Content is automatically indexed for real-time semantic search during conversations.

Multi-Channel Deployment

Deploy your agents to webchat, WhatsApp, Messenger, Instagram, and WeChat from a single project with channel-specific configurations.

Full API Access

Interact with your agents programmatically via a REST API. Send messages, manage sessions, search knowledge bases, and retrieve analytics.

Analytics & Transcripts

Track conversation metrics, resolution rates, and message volumes. Review full conversation transcripts to evaluate and improve agent performance.

How It Works

1

Create a Project

Start from scratch or use a template. Choose between the agentic framework for autonomous agents or conversation-flow for structured step-by-step dialogs.

2

Design Your Agent

Use the visual canvas to add steps (messages, prompts, conditions, API calls, KB searches), connect them, configure agent instructions, and upload knowledge base documents.

3

Test & Iterate

Preview your agent in the built-in webchat widget. Review transcripts, check analytics, and refine prompts and workflows until your agent performs as expected.

4

Publish & Deploy

Publish a version snapshot, then deploy to webchat (embeddable widget), WhatsApp, Messenger, Instagram, or WeChat. Use the API for custom integrations.

Get Started