Skip to main content
Version: Next

Overview

Obot is a complete platform for building and running AI agents. This guide will help you choose the right deployment method for your use case.

Obot Components​

Obot consists of three main components:

  • Obot Server: The core application server
  • PostgreSQL Database: Version 17 or higher with pgvector extension
  • Data Storage: Local filesystem or S3-compatible storage for workspace files

Obot stores its data under the /data path. It also includes the data for the built-in PostgreSQL instance during development. Production deployments require an external PostgreSQL database.

Deployment Options​

Docker Deployment​

Best for: Local development, testing, proof-of-concept

Docker provides the fastest way to get Obot running on your local machine or a single server.

  • Simple setup with docker run
  • Ideal for development and evaluation
  • Uses built-in PostgreSQL

👉 Docker Deployment Guide

Kubernetes Deployment​

Best for: Production deployments, scalability, high availability

Deploy Obot on Kubernetes for production-grade reliability and scalability.

  • Helm chart available at charts.obot.ai
  • Integrates with cloud services (KMS, S3, etc.)
  • Requires external PostgreSQL database

👉 Kubernetes Deployment Guide

Cloud Platform Reference Architectures​

Best for: Planning production deployments on cloud-managed Kubernetes

If you're planning to deploy Obot on cloud-managed Kubernetes services, these reference architectures provide infrastructure guidance and best practices.

  • Infrastructure blueprints: Pre-configured setups using cloud-native services
  • Best practices: Security, networking, and scalability recommendations
  • Managed services integration: Databases, storage, and key management

Reference architectures for cloud-managed Kubernetes:

System Requirements​

Minimum (Development/Testing)​

  • CPU: 1 cores
  • RAM: 2 GB
  • Storage: 10 GB

Database Requirements​

  • Development: Built-in PostgreSQL included
  • Production: External PostgreSQL 17+ required with pgvector extension

Production Considerations​

For production deployments, you should have:

  • External PostgreSQL database: PostgreSQL 17+ with pgvector extension
  • S3-compatible storage: For workspace files and data
  • Encryption provider: AWS KMS, Google Cloud KMS, or Azure Key Vault
  • Authentication: OAuth, OIDC, or enterprise providers (SAML, LDAP)
  • TLS/SSL certificates: For secure HTTPS access
  • Backup strategy: Regular backups of database and storage

Quick Decision Guide​

Use CaseRecommended Deployment
Local developmentDocker
ProductionKubernetes

Next Steps​

  1. Choose your deployment method above
  2. Follow the deployment guide
  3. Configure authentication
  4. Set up model providers
  5. Review server configuration

Getting Help​