|
|
--- |
|
|
title: Patent Architect AI v2 |
|
|
emoji: 🏗️ |
|
|
colorFrom: green |
|
|
colorTo: yellow |
|
|
sdk: gradio |
|
|
sdk_version: "4.44.0" |
|
|
app_file: app.py |
|
|
pinned: false |
|
|
--- |
|
|
|
|
|
# 🏗️ Patent Architect AI v2 - True Agentic Workflow |
|
|
|
|
|
**An AI-powered system that uses a team of collaborating agents to transform a raw idea into a strategized patent draft.** |
|
|
|
|
|
This application demonstrates a true multi-agent system where different AI agents, each with a specific role, negotiate and collaborate to produce a comprehensive patent application. It includes live prior art searches and multiple forms of AI-generated visual aids. |
|
|
|
|
|
## 🎯 Key Features |
|
|
|
|
|
### **🤖 True Agentic Collaboration** |
|
|
A team of specialized AI agents work together, with the output of one agent influencing the actions of the next. |
|
|
- **Prior Art Detective**: Performs *live* web searches on Google Patents and Wikipedia to find relevant prior art. |
|
|
- **Chief Strategist**: Analyzes the prior art and formulates a defensible patenting strategy. |
|
|
- **Technical Writer, Claims Drafter, Figure Drafter**: Generate the core sections of the patent, all adhering to the Chief Strategist's mandate. |
|
|
- **Conceptual Artist**: Generates a high-quality conceptual image of the invention using the Segmind Ideogram API. |
|
|
|
|
|
### **🎨 Dual Visual Outputs** |
|
|
The system generates two distinct, real images to aid understanding: |
|
|
- **Conceptual Image**: A photorealistic image from the Segmind API that helps visualize the invention's marketing potential. |
|
|
- **Technical Figure**: A patent-style drawing rendered from real LaTeX/TikZ code, which is essential for the patent document itself. |
|
|
|
|
|
### **💡 From First Principles** |
|
|
This application is built with a focus on robust, elegant design. It handles errors gracefully and makes agent interactions transparent, demonstrating a resilient and well-architected system. |
|
|
|
|
|
--- |
|
|
|
|
|
## 🚀 Getting Started |
|
|
|
|
|
### 1. Install Dependencies |
|
|
Ensure you have Python 3.8+ installed, then run: |
|
|
```bash |
|
|
pip install -r requirements.txt |
|
|
``` |
|
|
|
|
|
### 2. Set Up API Keys (Crucial Step) |
|
|
This application requires API keys to function. Your keys are not stored or shared. |
|
|
|
|
|
#### **For Local Development:** |
|
|
1. **Create a `.env` file** in the main project directory. |
|
|
2. **Copy the text below** into your `.env` file. |
|
|
3. **Add your secret keys** in place of the placeholder text. |
|
|
|
|
|
``` |
|
|
# .env file for Patent Architect AI v2 |
|
|
|
|
|
# Required for the core language model agents (Strategist, Writer, etc.) |
|
|
# Get a key from Google AI Studio: https://makersuite.google.com/app/apikey |
|
|
GEMINI_API_KEY="YOUR_GEMINI_API_KEY_HERE" |
|
|
|
|
|
# Required for the PriorArtDetective agent to perform live web searches. |
|
|
# Get a free key from SerpApi: https://serpapi.com/ |
|
|
SERPAPI_API_KEY="YOUR_SERPAPI_API_KEY_HERE" |
|
|
|
|
|
# Required for the ConceptualArtist agent to generate high-quality images. |
|
|
# Get a free key from Segmind: https://segmind.com/ |
|
|
SEGMIND_API_KEY="YOUR_SEGMIND_API_KEY_HERE" |
|
|
``` |
|
|
|
|
|
#### **For Hugging Face Spaces Deployment:** |
|
|
1. Do **not** upload your `.env` file. |
|
|
2. In your Space settings, go to the **"Secrets"** section. |
|
|
3. Add each of the three keys (`GEMINI_API_KEY`, `SERPAPI_API_KEY`, `SEGMIND_API_KEY`) as a separate secret. |
|
|
|
|
|
### 3. Run the Application |
|
|
Once your API keys are configured, run the app: |
|
|
``` |
