Skip to main content

​
Frequently Asked Questions

Nadoo AI is an open-source AI Agent Builder Platform for creating, deploying, and managing intelligent AI agents. It provides:
  • A visual workflow engine (powered by LangGraph) with 18+ node types and 6 agent strategies
  • A RAG-powered Knowledge Base with hybrid search (vector + BM25) and reranking
  • Support for 12+ AI model providers through a unified API
  • Multi-channel messaging to deploy agents across Slack, Discord, Telegram, KakaoTalk, Teams, and WhatsApp
  • An extensibility layer with the Plugin SDK and Skills system
Nadoo AI is designed for teams that need to build production-ready AI agents without vendor lock-in.
Yes. Nadoo AI is released under the BUSL-1.1 (Business Source License 1.1). It is free to use for development, testing, and internal business use. The license allows you to run the platform on your own infrastructure without licensing fees.For specific licensing terms regarding commercial redistribution, refer to the LICENSE file in the repository.
Nadoo AI supports 12+ AI model providers through a unified provider abstraction:
ProviderModels
OpenAIGPT-4o, GPT-4, GPT-3.5 Turbo
AnthropicClaude 4, Claude 3.5 Sonnet, Claude 3 Haiku
GoogleGemini 1.5 Pro, Gemini 1.5 Flash
Azure OpenAIEnterprise-hosted OpenAI models
AWS BedrockClaude, Titan, Mistral, and more
Google Vertex AIGemini models via Google Cloud
OllamaLlama, Mistral, Phi, and other local open-source models
vLLMHigh-throughput inference for any supported model
Custom endpointsAny OpenAI-compatible API
You can configure multiple providers per workspace and select the model at the node level in your workflows.
See the AI Models page for detailed configuration instructions for each provider.
It depends on what you want to do:
  • No coding required: The visual workflow builder lets you design AI agent logic by dragging and connecting nodes. You can build complete agents, configure knowledge bases, and deploy to messaging channels entirely through the web UI.
  • Coding required: Developing custom plugins with the Plugin SDK requires Python knowledge. Building custom skills requires writing SKILL.md manifests and supporting scripts.
The platform is designed to be accessible to both non-technical users and developers.
Nadoo AI requires two databases:
  • PostgreSQL 15+ with the pgvector extension — stores all application data and vector embeddings for the Knowledge Base
  • Redis 7+ — used for task queuing (Celery), caching, real-time messaging (WebSocket), and session management
Both are included in the Docker Compose configuration and start automatically when you run npm run start.
For production deployments, we recommend using managed database services (e.g., AWS RDS for PostgreSQL, Amazon ElastiCache for Redis) for reliability and automated backups.
Yes. Nadoo AI is designed for self-hosted deployment. The repository includes Docker Compose configurations for multiple environments:
  • Local development — single-machine setup with hot-reload
  • Development/staging — multi-container setup with production-like settings
  • Production — optimized configuration with security hardening, TLS, and resource limits
See the Self-Hosting Guide and Docker Compose Local for step-by-step deployment instructions.
The Knowledge Base supports ingestion of the following document formats:
  • PDF — including scanned documents with OCR
  • Microsoft Word (DOCX)
  • Plain text (TXT)
  • Markdown (MD)
  • Microsoft Excel (XLSX) — tabular data extraction
  • Web pages — URL-based scraping and ingestion
Documents are automatically chunked, embedded, and indexed for hybrid search (vector similarity + BM25 keyword matching). Optional reranking further improves retrieval relevance.
Custom data loaders can be built using the Plugin SDK to support additional formats.
Channels connect your AI agents to external messaging platforms. Each channel integration is handled by a dedicated adapter that manages authentication, message formatting, and bidirectional communication.Supported channels:
  • Slack — Bot integration with slash commands and thread support
  • Discord — Bot with channel and DM support
  • Telegram — Bot API integration
  • KakaoTalk — Channel and chatbot integration
  • Microsoft Teams — Bot Framework integration
  • WhatsApp — Business API integration
You configure channels at the workspace level. Each channel is linked to an application (AI agent), so messages received on that channel are routed to the corresponding workflow for processing and response.See the Channels Overview for setup instructions.
Yes. Nadoo AI provides a comprehensive REST API at /api/v1/ with 530+ endpoints covering:
  • Workspace and user management
  • Application (agent) CRUD and execution
  • Workflow creation and management
  • Knowledge base operations (upload, search, manage documents)
  • Chat and conversation history
  • Channel configuration
  • Model provider management
  • Plugin and skill administration
The API uses JWT-based authentication and supports both API key and OAuth2 access. Interactive API documentation is available at http://localhost:8000/api/docs (Swagger UI) when the backend is running.See the API Overview and Authentication pages for details.
Nadoo AI is an open-source project and contributions are welcome. Here is how to get involved:
  1. Read the contributing guide — See CONTRIBUTING.md in the repository for coding standards, PR process, and development setup
  2. Browse GitHub issues — Look for issues labeled good first issue or help wanted at github.com/nadoo-ai
  3. Join the community — Connect with other contributors on Discord in the #contributing channel
  4. Submit pull requests — Fork the repository, make your changes, and open a PR against the main branch
All contributions — code, documentation, bug reports, and feature suggestions — are appreciated.

​
Still Have Questions?

Discord Community

Join our Discord for real-time help and discussions

GitHub

Browse source code and open issues

Troubleshooting

Solutions to common technical issues

Glossary

Definitions of terms used throughout the platform