Zatoona
Enable non-technical users to query and analyze local or cloud datasets conversationally via AI-agent orchestration using open-source data-science tools.
Project Description
Background & Business Case:
Small-to-medium businesses (SMBs) as well as science research teams in universites are increasingly generating large volumes of valuable data but lack the internal expertise or budget to transform this data into actionable insights. Hiring data analysts or data science teams is costly, and existing tools require technical know-how. This creates a gap between the availability of data and the ability to leverage it for business decisions
We developed a natural language-driven AI assistant that allows non-technical users to interact with their data conversationally.
Users provide paths to their datasets (e.g. from sales, marketing, or operations) either locally on the cloud. They could then inteact with their data, ask questions in plain English, and receive analytical insights.
This made possible powered by an orchestration of AI agents using open-source data science tools on the data either on the cloud or on the users local infra.
To enable this the we the user setup the project, a docker enviroment is created with all the required dependecies. User queries are then converted to code that is evaluated and reviewed and then excuted in the enviroment. Results are then returned to the user as a natural language, visualizations, reports, etc.