Agentic Image Flow (LangGraph style)

This repo contains a minimal, local LangGraph-like SDK and an example agentic flow that: 1) creates content from a prompt, 2) generates an image from the content, 3) post-processes that image (resize + watermark), and 4) generates concise alt text.

Quick start

1. Install dependencies:

pip install -r requirements.txt

2. Set your OpenAI API key in environment:

Windows (PowerShell):

$env:OPENAI_API_KEY = "sk-..."

Linux / macOS:

export OPENAI_API_KEY="sk-..."

3. Run the flow:

python run_flow.py --prompt "A vintage poster of a robot baker" --outdir output

Outputs will be written into the output directory.

Notes

• This is a minimal example to demonstrate an agentic flow. Replace the OpenAI calls or extend nodes as needed.
• If you want to swap in a different image-generation provider, modify flows/agentic_image_flow.py in the generate_image node.

🏥 Medical Appointment System using LangGraph, FastAPI, and Streamlit

A multi-agent, AI-powered Doctor Appointment Booking System designed to handle user queries about doctor availability, specialization, and appointment scheduling.
This project demonstrates intelligent workflow automation between agents using LangGraph, LangChain, and a clean FastAPI–Streamlit integration.

---

🚀 Features

• AI-driven appointment scheduling and doctor recommendations
• Multi-agent coordination for query handling and decision-making
• Dynamic workflow automation using LangGraph
• Simple, interactive frontend built with Streamlit
• REST API powered by FastAPI for backend execution
• CSV-based data management with Python and Pandas

---

🧠 Tech Stack

Technology	Purpose
LangGraph	Workflow automation between agents
LangChain	Model loading, prompt creation, and tool usage
FastAPI	Serves API endpoints and executes logic
Streamlit	Frontend interface for user interaction
Python + Pandas + CSV	Data handling and storage

---

🧩 Architecture Overview

1. User Input → Streamlit UI
2. Query Handling → LangChain-powered Agent
3. Workflow Coordination → LangGraph automates inter-agent communication
4. Backend Execution → FastAPI processes requests and returns responses
5. Data Management → Pandas reads/writes appointment data via CSV

---

🩺 Example Use Case

---

⚙️ How to Run Locally

# Clone the repository
git clone https://github.com/yourusername/medical-appointment-system.git

# Navigate to the project directory
cd medical-appointment-system

# Install dependencies
pip install -r requirements.txt

# Run the FastAPI backend
uvicorn main:app --reload

# Run the Streamlit frontend
streamlit run app.py