A powerful Multi-Agent System built with a modular architecture to perform:
- Video & Web summarization
- Conversational search
- AI-powered image generation
This Streamlit-based app coordinates multiple agents to communicate with various external APIs and services, including LLMs, Wikipedia, Arxiv, and DuckDuckGo β enabling intelligent, user-facing applications.
- Video & Web Summarizer: Uses Groq LLMs and web/YT scrapers
- Chat with Search: Conversational agent with knowledge access
- Image Generator: Generates contextual images from user prompts
- Groq LLM Service β Fast LLM-based reasoning and summarization
- Wikipedia API β Structured encyclopedia lookup
- Arxiv API β Scientific paper access
- DuckDuckGo Search β General web search
- Flux API β Image generation support
- YouTube Transcript Loader β Extracts captions for analysis
- Webpage Scraper β Summarizes article content
- Built-in API Key Store: Dynamically injects API keys into agents at runtime
- Frontend: Streamlit
- Backend: Python multi-agent architecture
- APIs: LLMs, Wikipedia, Arxiv, YouTube, DuckDuckGo, Flux
- Browser: Runs client-side in any modern browser
git clone https://github.com/your-repo/multi-agent-app.git
cd multi-agent-app