amcrisan / dataagent

Flask based REST API for experimenting with multi-agent systems that support data analysis and visualization

Data Agent

This is a flask-based API to support experimentation with single and multi-agent systems that support data analysis and visualization. Data Agent advances a framework for modeling AI and Human agents collaborating through a shared environment. This library also provides lightweight support for applications relying on conversationg interfaces.

Getting Started

Ensure that you have Python 3.12 installed. This project uses pipenv to manage dependencies. You can install pipenv as follows: pip install pipenv

Dependencies

Downloading, manging, and otherwise modifying dependencies can be managed using pipenv install, much like you would use pip. It is also possible to directly modify the Pipfile. Dev dependencies, for example jupyter notebook and other tools that are useful for exploration, but not the core funcitonality of the library, can be installed using the pipenv install --dev option. Please consult pipenv itself for additional commands and options.

Running the API

Data Agent is a flask-based REST API, intended to be accessed by other applications.

Set the environment variable so that the flask app points to ./DataAgent/api.py, this way you can easily run the api from the parent directory. export FLASK_APP=/DataAgent/api.py. The config.py file is automatically set into development mode.

Usage

As this is a library intended to support experimentaion it is intended to support varied functionality. As a result, the maturity of different functional capabilities will also vary. As some ideas solidfy they may become their own separate libraries dedicated to a single purpose.

Refer to following for additional details:

• Routing : An overview of the different routes this API supports and that can be called by client facing applications.

• Models: An overview of the models (Human, AI Agent, Environment) that are currently supported and how to use them. Models that support applications (e.g., Conversational Agents) can also be found here

• Examples: Some general examples of use -- includes Jupyter notebooks. Ideas and future looking experiments can also be found here.