File size: 3,289 Bytes
52d7f2a |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 |
---
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:
``` |