Gurubase home page
Search...
⌘K
Gurubase.io
Gurubase.io
Search...
Navigation
Get Started
Quickstart
Documentation
API Reference
Try Gurubase.io
Get Started
Welcome to Gurubase
Quickstart
Self-Hosted (Onprem)
Guides
Create Your First Guru
Analyze Usage
Prompt Your Guru
PII Masking
Excel Extraction Best Practices
Integrations
Bots
Ingestion
On this page
Getting Started
Self-hosted Enterprise Edition
Architecture Overview
Additional Resources
Next Steps
Get Started
Quickstart
Get started with Gurubase in minutes
Getting Started
You can use Gurubase through our cloud platform:
Gurubase.io
- Start using immediately without any setup
Self-hosted Enterprise Edition
Self-hosted installations are now exclusively available under the Enterprise plan. To request access to the self-hosted package, please contact us at
[email protected]
.
Architecture Overview
Gurubase follows a microservices architecture designed for scalability and performance. Here’s a quick overview of the key components:
Core Components
Frontend
: Next.js application for the user interface
Backend
: Django REST framework for API and business logic
Vector Store
: Milvus for similarity search and embeddings
Message Queue
: RabbitMQ for task distribution
Cache
: Redis for rate limiting and session management
Database
: PostgreSQL for application data
Key Features
RAG System
: Retrieval Augmented Generation for accurate answers
Multi-source Processing
: Support for websites, PDFs, YouTube videos and GitHub repositories
Embedding Options
: OpenAI or custom embedding API
Trust Scoring
: LLM-based context relevance evaluation
Binge Sessions
: Visualize your learning path while talking with a Guru. You can navigate through it and create a personalized path
Additional Resources
For detailed information about:
Advanced configuration
System architecture
Available services
Data storage
Security considerations
Upgrade and removal instructions
Next Steps
Create Your First Guru
Learn how to create and configure your first AI assistant
Welcome to Gurubase
Self-Hosted (Onprem)
Assistant
Responses are generated using AI and may contain mistakes.