Description
Hexabot V3 offers a self-hosted AI workflow automation platform designed for building AI agents, agentic workflows, and business automation with complete control. Users can self-host Hexabot, connect various channels, orchestrate tools and actions, and leverage memory and knowledge bases while keeping production data securely under their control.
The platform facilitates building workflows locally using a CLI tool. Installation involves installing the Hexabot CLI, creating a project from a starter template, navigating into the project workspace, and running the local Hexabot stack. This setup allows developers to build workflows that integrate conversations with actions, channels, and memory.
Key features include self-hostable AI workflow automation for real operations, bringing conversations, actions, memory, governance, and multichannel service delivery into a single runtime. Workflows can be triggered by chat, forms, schedules, or events, and can call models, tools, APIs, and back-office systems. Model flows can be controlled visually or via portable YAML with typed contracts, schema validation, and version control. Actions and bindings allow for the connection of reusable actions, LLM calls, capability bindings, and external systems with defined input, output, and settings contracts.
Hexabot supports managed memory, retrieval-augmented context (RAG), and MCP integration points for sanctioned tools. Human review points can be implemented using conditionals, loops, segmentation, inbox workflows, and live handoff for scenarios requiring human judgment. The platform is ready for service operations, enabling multilingual assistants across web, messaging, social, and custom channels with features for roles, analytics, and team control.
A success story with Orange demonstrates Hexabot's capability in running multilingual service automation across website and social channels, effectively handling frequent requests, routing complex cases, and maintaining a consistent customer experience. Pricing models cater to different needs, from a free community plan for early prototypes to paid tiers like Starter, Pro, and Scale for production use, with a custom Enterprise option for tailored solutions.
Hexabot V3 is source-available under the Fair Core License, Version 1.0, ALv2 Future License, allowing inspection, self-hosting, and extension of the codebase. Commercial license-key plans unlock higher capacity for users, workflows, and activations, with deployment options on-premise or in the cloud, ensuring full source visibility and control over production data, workflow logic, memory, and extensions. Users can share components like actions, channels, and helpers with the community.
Hexabot's Core Features
Self-hostable AI workflow automation
Build AI agents and agentic workflows
Connect any channel (web, messaging, social, custom)
Orchestrate tools, actions, and APIs
Managed memory and RAG capabilities
Visual and YAML workflow control
Typed contracts and schema validation
Human review points and live handoff
Multilingual assistant support
Service operations readiness
Source-available under Fair Core License
Control over production data and infrastructure
How to use Hexabot?
Install CLI: Use npm to install the official Hexabot CLI.
Create Project: Use the CLI to scaffold a self-hostable project from a starter template.
Navigate Workspace: Move into your newly created project directory.
Run Local Stack: Execute the command to start the local Hexabot development environment.
Define Workflows: Configure workflows, connect channels, and extend runtime with actions and tools.
Connect Channels: Integrate with various communication platforms and entry points.
Extend Runtime: Add custom actions, tools, and bindings to enhance capabilities.
Deploy & Operate: Run and manage your AI workflows for real-world service operations.
Hexabot's Use Cases
- Customer Service Automation
- Business Process Automation
- Multilingual Assistants
- Data Control Workflows
- Tool Orchestration
- Prototyping AI Agents
- Service Delivery Automation
