The Research Agent is an AI-powered tool designed to assist users in conducting research by searching for information, finding relevant articles, and summarizing them automatically. This tool utilizes advanced natural language processing (NLP) techniques and web scraping to provide users with concise summaries of relevant articles on their chosen topics.

Users can enter their research query into the search bar provided in the Streamlit app. The Research Agent then utilizes the Google SERPER API to perform a search query and retrieve relevant articles based on the user's input.

Once the search results are obtained, the Research Agent displays a list of relevant articles along with their titles and URLs. Additionally, it uses the Linkpreview library to fetch a preview of each article, including the title, description, and thumbnail image (if available), providing users with a quick overview of each article's content.

After selecting an article from the list, users can click on the URL to view the full article. The Research Agent then uses Selenium to scrape the text content of the article from the webpage. Once the text content is obtained, it applies recursive character text splitting to break down the article into smaller chunks suitable for processing.

Using Langchain and OpenAI's GPT-3.5 model, the Research Agent generates a concise summary of the rticle's content. This summary is presented to the user within the Streamlit app, allowing them to quickly grasp the main points of the article without having to read through the entire text.

1. Clone the repository:

git clone https://github.com/Hk669/research-agent.git
cd research-agent

2. Install dependencies:

pip install -r requirements.txt

3. Set up environment variables:

Create a .env file in the root directory and add your SERPER and OpenAI API keys:

SERPER=your_serper_api_key
API_KEY=your_openai_api_key

1. Run the Streamlit app:

streamlit run agent.py

2. Enter your query in the text input field and click the "Search" button.

3. The app will display a list of relevant articles along with their titles and URLs.

4. Click on the URL to view the full article.

5. The app will generate a concise summary of the articles.

• Python
• Streamlit
• Selenium
• Langchain
• Linkpreview

This project is licensed under the MIT License.