Goal

Provide lobbyist with initial positions of senator with regards to certain policies, laws or trending ideas as a first step to implement an approaching sponsoring-strategy.

How it works

  • Using Google's Gemini AI to assess senatorial stance on various policies, laws and ideas based on their public statements on X.com (Twitter.com).

  • The interaction is made through a Streamlit Application which results into:

    • A Table with all senators twitter handles separated into Support, Opposition, Neutral and Undecided with regard to the asked question by user.
    • A Pie Chart to visualize, in percentage points, the distribution of senators' positions.

Challenges we ran into

  • Customized the question into a suitable prompt.

  • Formatting the response into an analysis ready data frame.

  • Integrate the solution into a user friendly, Streamlit web application.

Accomplishments that we're proud of

Turning the tedious task of analyzing thousands of tweets into one click action.

What's next for Lobby{ai}st Buddy

This exercise a based only on one type of data (tweets data), for more stable results, it is encouraged to incorporate:

* Voting histories of each senators (both in local and federal)
* Complete history statements in multiple social media plate-forme
* Relationship with pressure groups, caucus...

Built With

  • generativeai
  • langchain
  • python
  • streamlit
Share this project:

Updates