Skip to content

Installation Guide

System Setup

PowerShell Installation

Install PowerShell using the Windows Package Manager (Winget):

winget install --id Microsoft.PowerShell --source winget
Verify installation PowerShell 7+:

pwsh --version

Install PowerShell using Homebrew:

brew install --cask powershell

Verify installation PowerShell 7.4.5:

pwsh --version

Tip

Please run all command using the latest version of PowerShell (pwsh) to ensure cross-platform compatibility.

Container Setup (Podman)

To install Podman on Windows, follow these steps:

  1. Download the Podman installer from the official website: Podman Windows Installer.
  2. Run the installer and follow the on-screen instructions to complete the setup.
  3. After installation, run below command in your terminal:
podman machine init

Install Podman using Homebrew:

brew install podman

After installation, ensure Podman is added to your PATH by verifying it:

podman --version
podman machine init

Repository Setup

Clone the repository from GitHub:

git clone https://github.com/Insight-Services-APAC/Insight_Ingenious.git

Navigate to the project directory and get the latest release:

cd Insight_Ingenious

git checkout v0.0.1

Profile Setup

  1. Please set up your OpenAI API by visiting https://oai.azure.com/ and developer_guide

  2. Create Your profiles.yml:

Use PowerShell to create or edit profiles.yml:

# Create or edit profiles.yml using your default code editor
# Follow the template at /conversation_pattern_example/profile.yml
code $home\.ingenious\profiles.yml

Set up the path for profiles.yml:

export INGENIOUS_PROFILE="$HOME/.ingenious/profiles.yml"

# Create or edit profiles.yml using a preferred text editor (e.g., nano, vim, or code)
# Follow the template at /conversation_pattern_example/profile.yml
code $INGENIOUS_PROFILE

Build the Docker Image

There are two methods to build and run the Docker image:

Run the installation script using PowerShell:

pwsh .\run_container_installation_dev.ps1
or 

pwsh .\run_container_installation_prod.ps1
if want deploy with a client site.

Tip

The default parameter for the installer is:

    param (
    [string]$podman_path = "C:\Program Files\Podman\bin\podman.exe",
    [string]$image_name = "localhost/ingen_dev_ubuntu",
    [string]$container_name = "ingen_dev_ubuntu",
    [string]$dockerfile_path = "./docker/development_images/linux_development_image_ubuntu.dockerfile"
you can modify it for difference images and base container.

Build the development Docker image using Podman:

podman build -f ./docker/linux_development_image.dockerfile -t localhost/ingen_dev2 ./docker/

After building the image, run the container using the PowerShell script:

pwsh .\run_image_in_podman.ps1

Verify Running Containers

Check the status of your containers to ensure everything is running correctly:

podman ps -a
This command displays a list of all running and stopped containers.

Access the Container (Optional)

If you use the simple Simple SH deployment, the script will automatically SSH into the container. The following steps will be optional.

To use Podman with the "Dev Containers" extension in Visual Studio Code:

  1. Open VSCode and navigate to Settings:
  2. Press Ctrl + , or Cmd + , (on macOS) to open settings.
  3. Search for Dev Containers: Docker Path.

  4. Change the path from docker to podman.

  5. Open the Command Palette (Ctrl + Shift + P or Cmd + Shift + P on macOS).

  6. Type Dev Containers: Attach to Running Container... and select your running Podman container.

This configuration enables you to use Podman as the container runtime within VSCode's "Dev Containers" extension.

You can access the running container directly via the command line:

podman exec -it <container-id> /bin/bash

Replace <container-id> with the ID or name of your running container.

Run Tests

Now you can test your container setup:

python run_ingen_cli.py

Open your browser and navigate to, to authenticate, click Authorize, and using Web Configuration credentials in profile.yaml.

http://localhost:9000/docs

Click on POST, then select Try it Out, and enter the following request for testing: 

{
  "thread_id": "sample",
  "user_prompt": "Tell me about basketball?",
  "user_id": "testuser123",
  "user_name": "testuser",
  "topic": "basketball, tennis, soccer",
  "memory_record": true,
  "conversation_flow": "classification_agent"
}
Please refer to service prefabs for available conversation flows.

In one Terminal:

While your session from run_ingen_cli.py is still active, OPEN another Terminal and SSH into the container:

python run_tests_http.py

Follow how to use guide for your first conversation in Python.