Adult Income Predictor & QueryBot

What it does

My app is designed for Knowledge Representation and Insights Generation from Structured Datasets. It streamlines data analysis by offering tools for preprocessing, feature engineering, and visualization. The app uses a logistic regression model for income prediction due to its high accuracy and provides an intuitive interface for data overview and visualization.

A key feature of the app is its query bot, initially powered by Google PaLM2 API integrated with LangChain. This setup allowed users to interact with the data through natural language queries, translating them into actionable data operations. However, after PaLM2 was deprecated, so I had to switch to using the Gemini API (model = gemini-pro). The transition maintained the app’s ability to handle natural language queries effectively.

With PandasAI interpreting the queries and Gemini executing the data operations, users can still generate downloadable CSVs or visualizations from their queries. This query bot enhances the app's accessibility, enabling users to derive insights from their data without needing advanced technical skills. The seamless integration of these advanced AI tools ensures the app remains both powerful and user-friendly, providing an efficient way to convert natural language into actionable data-driven insights.