A modular, multi-agent control panel built with Streamlit. This application acts as an orchestrator that routes user tasks to four specialized backend AI agents: Web Search, Code Execution, Document Parsing, and Browser Automation.
This project demonstrates a clean, decoupled architecture for managing specialized AI agents. Using a centralized Streamlit user interface, users can seamlessly switch between different operational modes depending on their current workflow requirements.
- Web Agent: Queries live internet data and returns structured title and link references.
- Code Agent: Accepts, executes, and returns outputs for dynamic Python code blocks.
- Document Agent: Ingests multiple file formats (
.pdf,.docx,.xlsx) and extracts structured text. - Browser Agent: Simulates programmatic, headless web browser actions and workflows.
- Isolated Architecture: Each agent operates out of an independent module under the
agents/directory for easy scaling and maintenance.
streamlit, multi-agent-system, ai-agents, web-scraping, code-interpreter, document-parsing, browser-automation, python, modular-architecture, llm-orchestration