Data Visualisation Dashboard Agent

Creating an AI Agent using openAI model and dash Package in python to create data visualisation dashboard for any given structured data.

Features

We have two agents in this project :

• Dashboard Analytics Agent: Processes datasets and provides insightful visualizations using GPT-based suggestions.
• Frontend Client: Allows users to interact with the analytics agent by sending datasets and fetching dashboard links.

-Visualization Generator: Automatically generates visualizations such as scatter plots and pie charts for submitted datasets.

Prerequisites

• Python 3.11+
• Virtual Environment: Recommended for managing dependencies.
• API Keys:
  • (OpenAI API Key)[https://platform.openai.com/settings/profile/api-keys]
  • (Fetch.ai Agentverse API Key)[https://agentverse.ai/profile/api-keys]
  • Identity seed phrases for agents (DASHBOARD_AGENT_KEY and CLIENT_KEY), random strings to get the same agent addresses everytime.

Installation

1. Clone the Repository

git clone https://github.com/abhifetch/dashboard-agents.git
cd dashboard_agents

2. Create a Virtual Environment

python3 -m venv venv
source venv/bin/activate  # For Linux/Mac
venv\Scripts\activate     # For Windows

3. Install Dependencies

pip install -r requirements.txt

Environment Variables

Create a .env file in the root directory of the project (where data_analytics_client.py and data_analytics_user.py are located).

Example .env File

OPENAI_API_KEY="your-openai-api-key"
AGENTVERSE_API_KEY="your-agentverse-api-key"
DASHBOARD_AGENT_KEY="your-dashboard-agent-seed-phrase"
CLIENT_KEY="your-client-agent-seed-phrase"

Running the Project

1. Start the Analytics Agent

Run the data_analytics_client.py to start the dashboard analytics agent:

python3 data_analytics_client.py

This will register the analytics agent with Fetch.ai's Agentverse and expose the webhook endpoint.

2. Start the Frontend Client

Run the data_analytics_user.py to start the frontend client:

python3 data_analytics_user.py

This will register the user client and enable interactions with the analytics agent.

How to Use

1. Search for Available Agents

Use the following curl command to search for registered analytics agents:

curl -X GET http://localhost:5002/api/search-agents

2. Send Dataset to Analytics Agent

Send a dataset to the analytics agent for visualization:

curl -X POST http://localhost:5002/api/send-data \
-H "Content-Type: application/json" \
-d '{
  "dataPath": "data.csv", # save the data with in the same directory as agent scripts. as it is taking it locally at the momement.
  "agentAddress": "replace-with-agent-address-from-search"
}'

3. Get Dashboard URL

Poll for the generated dashboard URL:

curl -X GET http://localhost:5002/api/get-dashboard-response

If successful, you will receive a URL for the visualizations dashboard.

Sample output

Sample output for this can be visited on : Youtube Video

If you wish to learn more about fetch.ai SDK please visit GitHub