Home Applications IRIS AI Studio

IRIS AI Studio

This application is not supported by InterSystems Corporation. Please be notified that you use it at your own risk.
0 reviews
IPM installs
Pull requests
This app has a demo View
AI Studio to load and retrieve vector embeddings from your files

What's new in this version

  1. Online Demo of the application is now available at https://iris-ai-studio.vercel.app
  2. Added Docker for easy local installation
  3. Connectors module can now load data from (OpenAI/Cohere embeddings)
    1. Local Storage
    2. AWS S3
    3. Azure Blob Storage
    4. Airtable
  4. Playground module is fully functional with
    1. Semantic Search
    2. Chat with Docs
    3. Recommendation Engine
      1. Cohere Re-rank
      2. OpenAI Re-rank
    4. Similarity Engine

IRIS AI Studio


A no-code/low-code tool to explore the capabilities of vector embeddings in InterSystems IRIS DB.

  • Connectors let load data from files as vector embeddings into IRIS DB

  • Playground let users explore different retrieval channels on the vector embeddings reside in IRIS DB

Process Flow

Try Online


Designed and developed only for web interface, not compatible with mobile

Step 1: Setup the Instance details and API Keys in settings

Step 2: Through connectors load data into IRIS DB

Step 3: In playground, explore different retrieval options

⚠️ This application is deployed on render.com and the IRIS instance & API key info used on the live version may have been logged. Strongly recommended to use only development or temporary IRIS instances to explore the solution, and deactivate or delete the keys after use. Also, backend runs on a tiny server and won’t be able to handle heavy workloads.

Tech Stack

  • Frontend: VueJS, TailwindCSS, Flowbite

  • Backend: Python, Flask

  • Database: InterSystems IRIS

  • Frameworks/Libraries/Services: Llama-Index, SQLalchemy-iris, OpenAI, Cohere

  • Infrastructure: Vercel (frontend hosting), Render (backend hosting)

Instructions to Run

Simply execute the script using the following command.


If any permission issue while executing the script, allow it through chmod +x build.sh

Access the UI at http://localhost:5173

Access the APIs at

Credentials for Local InterSystems IRIS Instance
username: demo
password: demo
hostname: localhost
port: 1972
namespace: USER

IRIS Instance

Follow the instructions to spin off a Cloud InterSystems Community Edition

Once you have the credentials, in the frontend application’s settings page the credentials can be added. You may add more than one instance and choose to use whichever one for data ingestion or retrieval process independently.

You may follow the following instructions to individually run the frontend and backend

Frontend (VueJS)

Start from application’s root directory

cd frontend
npm i
npm run dev

Backend (Python)

Start from application’s root directory

cd backend
pip install -r requirements.txt
gunicorn app:app

Access the APIs at

Folder Structure

│   ├── src/
│   ├── .env                 
│   └── ...
├── backend/
│   └── app.py
│   ├── data_loader.py
│   ├── chat_llama.py
│   ├── query_llama.py
│   ├── similarity_llama.py
│   ├── reco_llama.py
│   ├── requirements.txt
│   │   └── ...
├── assets/
├── README.md
└── https://github.com/ikram-shah/iris-ai-studio/blob/main/LICENSE


This project is licensed under the MIT License.

You can find the full text of the license in the https://github.com/ikram-shah/iris-ai-studio/blob/main/LICENSE file.

Read more
Made with
1.0.116 May, 2024
Works with
InterSystems IRISInterSystems IRIS for HealthInterSystems Vector SearchInterSystems IRIS Cloud SQL
First published
11 May, 2024