Multi-agent research assistant that takes a topic, searches the web, scrapes sources, and generates a polished report with critic feedback — all through a simple Streamlit interface.

Research-Pilot is a multi-agent research application built with Streamlit and LangChain that automates the end-to-end research workflow. Given a topic, it searches for recent information via the Tavily API, scrapes the most relevant source with BeautifulSoup, drafts a comprehensive report using Groq or Mistral models, and runs a critic feedback loop to refine the output. Results are presented in a clean UI with expandable sections for search results, scraped content, the final report, and critic feedback — all downloadable as Markdown.