Skip to main content

Self Hosted

Architecture Diagram​

Overview of Panto architecture diagram

architecture diagram

We’ll help you deploy the Panto AI Agent (server) to your own cloud environment.

Prerequisites​

1. Version Control System​

We currently support the following providers:

  • GitHub
  • GitLab
  • Bitbucket
  • Azure DevOps

2. Deployment Environment​

Any compute platform capable of running Docker images is supported.
Examples: AWS ECS, Kubernetes (K8s), etc.

3. LLM Configuration​

You must configure access to at least one supported LLM:

  • OpenAI
  • Claude
  • Gemini
  • DeepSeek

You can either provide your own API keys or request access via our managed keys.

4. Database​

  • One PostgreSQL instance is required.

5. Load Balancer / API Gateway with SSL​

Required only if you're using a third-party version control system (e.g. github.com, gitlab.com).
This ensures secure webhook delivery to your server.

6. Observability​

  • You can export container logs to your preferred observability stack (e.g. Elasticsearch, Datadog, etc.).

7. Version Management​

  • We’ll notify you of updates along with new Docker image URLs and upgrade instructions.
  • Your deployment must remain within 30 days of the latest available version.

Deployment Instructions​

  1. We will share:

    • Docker image URLs with credentials
    • Required environment variables

  2. Deploy the Docker container using your preferred orchestration system.

  3. Verify the deployment by running the following command:

    curl -v http://<your_local_ip>:5001/health

  4. Once the health check passes, follow the provided installation documentation.

    • Note: Instead of our webhook URL, us your deployed server url as webhook.

Support​

For any support, drop us an email at hello@getpanto.ai