Inspiration

The goal of enabling common people and non-experts to interact with government spending data can serve as a significant source of inspiration. LLMs can be used to democratize data access by translating technical legalese or opaque financial concepts into understandable English explanations when combined with the available datasets.

What it does

Our project leverages cutting-edge AI to analyze data from USAspending.gov. You can search for spending by agency, year, or keyword, and get easy-to-understand summaries and visualizations. We're making government spending more transparent for everyone!

How we built it

We started by pulling data from USAspending.gov and structuring it in a way that our AI model could understand. We then used the Gemini API to analyze this data and generate insightful summaries. Finally, we built an interactive web application using Streamlit to allow users to easily access these insights. To make the data easy to explore, we used Tableau to build dynamic and interactive dashboards. The visualizations included graphs, charts, and filters, allowing users to explore different aspects of federal spending based on their interests.

Challenges we ran into

Handling Ambiguity in Data: Some of the spending data lacked context or clarity, making it challenging for the LLMs to generate precise insights. We had to fine-tune the models to ensure they provided useful and accurate information

Accomplishments that we're proud of

In just 36 hours, we gone through tons of USASpending.gov data and analyzed it. We figured out how to get the data from USAspending.gov and organize it so the AI could understand it and used tableau to visualize it.

What we learned

Federal spending data is often dense and difficult for the public to understand. Pre-trained LLMs can simplify this information by extracting and summarizing key insights. Tools like Tableau help transform raw data into clear, interactive visuals, making complex financial details easier to digest and analyze. While USAspending.gov provides valuable data, combining it with AI augments the insights and fills in gaps that make the information more meaningful for the public.

What's next for Smart Budget Bot

We intend to provide voice recommendations for real time experience. With the use of the usaspending.gov api-endpoints, we can verify the data from pre-trained LLMs.

Enhanced visualizations: Incorporate charts and graphs to display spending patterns. Agency-specific analysis: Provide deeper insights into individual agencies' spending. Comparative analysis: Allow users to compare spending across different agencies or years. Data updates: Implement automatic data updates from USAspending.gov.

Built With

Share this project:

Updates