GraceNgina-sudo/multi-agent-workflow-optimizer

workflow Automation for small businesses in Africa # Multi-Agent Workflow Optimizer An experimental multi-agent system built in python that simulates how specialized agents can plan, optimize, and execute tasks collaboratively. This project started as a leraning exercise and has since grown into a more strucutured application with a backened (FastAPI), database logging, and a frontend interface for user interaction ## Purpose The project's core purpose is to explore multi-agent systems in a practical way while learning modern development practices. ### Key goals: - Understand object-oriented programming (oop) and agent-based design in python - Simulate collaborative workflows between different agents (planner, Optimizer, Executor, Coordinator). - Gain hands-on experience with FastAPI, databases, and frontend-backend integration. - Practice software engineering principles: logging, modularity, and system design. - Lay the foundation for future integration with AI/LLMs for smarter decision-making. ## What it does: - Accepts a user query via the frontend (form input). - Planner Agent: Interprets the query and creates a task plan. - Optimizer Agent: Refines the plan for efficiency. - Executor Agent: Simulates carrying out the task. - Coordinator Agent: Logs the full workflow and maintains history. - All actions are logged into a SQLite database for traceability. - Users can view task history and workflow logs via API endpoints. ## Tech stack: - Backend: FastAPI (Python) - Frontend: JS + Bootstrap (form-based interface) - Database: SQLite (lightweight logging & workflow storage) - Server: Uvicorn (ASGI server for FastAPI) - Architecture: Modular agent-based design with orchestrator ## Usage 1. Run the FastAPI app: uvicorn main:app --reload 2. Open the frontend (with form connected to /ai-simulate) 3. Enter a query 4. Agents collaborate to: Plan-optimize-execute-log 5. View results in the UI or query logs via endpoints ## Current Features - Multi-agent collaboration (Planner, Optimizer, Executor, Coordin …