changed the title section to be more visible to the amd team
Browse files
README.md
CHANGED
|
@@ -1,15 +1,76 @@
|
|
| 1 |
---
|
| 2 |
-
title:
|
| 3 |
-
emoji:
|
| 4 |
-
colorFrom:
|
| 5 |
-
colorTo:
|
| 6 |
sdk: gradio
|
| 7 |
sdk_version: 6.14.0
|
| 8 |
-
python_version: '3.13'
|
| 9 |
app_file: app.py
|
|
|
|
| 10 |
pinned: false
|
| 11 |
license: mit
|
| 12 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
---
|
| 14 |
|
| 15 |
-
|
|
|
|
| 1 |
---
|
| 2 |
+
title: AMD HuggingFace Demo
|
| 3 |
+
emoji: 🚀
|
| 4 |
+
colorFrom: blue
|
| 5 |
+
colorTo: purple
|
| 6 |
sdk: gradio
|
| 7 |
sdk_version: 6.14.0
|
|
|
|
| 8 |
app_file: app.py
|
| 9 |
+
python_version: '3.13'
|
| 10 |
pinned: false
|
| 11 |
license: mit
|
| 12 |
+
tags:
|
| 13 |
+
- ai-agents
|
| 14 |
+
- startup-ideas
|
| 15 |
+
- langgraph
|
| 16 |
+
- multi-agent
|
| 17 |
+
- paul-graham
|
| 18 |
+
- amd-hackathon
|
| 19 |
+
short_description: Autonomous multi-agent system for startup idea discovery
|
| 20 |
+
---
|
| 21 |
+
|
| 22 |
+
# VentureForge 🚀
|
| 23 |
+
|
| 24 |
+
An autonomous multi-agent system that discovers startup ideas by mining real user pain points from online communities, evaluating them with Paul Graham-inspired criteria, and generating investor-ready pitch briefs.
|
| 25 |
+
|
| 26 |
+
**Built for AMD AI Hackathon | Track 1: AI Agents & Agentic Workflows**
|
| 27 |
+
|
| 28 |
+
## Features
|
| 29 |
+
|
| 30 |
+
- 🔍 **Pain Point Mining** from Hacker News, Product Hunt, and web search
|
| 31 |
+
- 🤖 **6 Specialized Agents** (Orchestrator, Miner, Generator, Scorer, Writer, Critic)
|
| 32 |
+
- ✅ **Binary Rubric Evaluation** (yes/no checks, not subjective scores)
|
| 33 |
+
- 🔄 **Reflection Loop** with adversarial critique (max 3 revisions)
|
| 34 |
+
- 📊 **Real-time Progress** tracking in Gradio UI
|
| 35 |
+
- 🎯 **Token Optimized** for vLLM 2048 token limit
|
| 36 |
+
|
| 37 |
+
## How to Use
|
| 38 |
+
|
| 39 |
+
1. **Set API Keys** in Space Settings → Repository secrets:
|
| 40 |
+
- `OPENROUTER_API_KEY` (required)
|
| 41 |
+
- `TAVILY_API_KEY` (required)
|
| 42 |
+
- `PRODUCTHUNT_API_KEY` (optional)
|
| 43 |
+
|
| 44 |
+
2. **Enter a domain** (e.g., "developer tools", "healthcare", "fintech")
|
| 45 |
+
|
| 46 |
+
3. **Click "Start Discovery"** and watch the agents work!
|
| 47 |
+
|
| 48 |
+
4. **View results** in the tabs:
|
| 49 |
+
- Pain Points (with evidence sources)
|
| 50 |
+
- Ideas (generated from pain points)
|
| 51 |
+
- Scored Ideas (8 binary checks)
|
| 52 |
+
- Pitch Briefs (investor-ready)
|
| 53 |
+
- Critiques (adversarial review)
|
| 54 |
+
|
| 55 |
+
## Architecture
|
| 56 |
+
|
| 57 |
+
```
|
| 58 |
+
Orchestrator → Pain Point Miner → Idea Generator → Scorer → Pitch Writer → Critic
|
| 59 |
+
↓
|
| 60 |
+
(reflection loop)
|
| 61 |
+
```
|
| 62 |
+
|
| 63 |
+
|
| 64 |
+
## Performance
|
| 65 |
+
|
| 66 |
+
- **Duration:** ~2 minutes per domain
|
| 67 |
+
- **Token Optimized:** One-at-a-time generation + compressed prompts
|
| 68 |
+
- **Reliable:** 0 JSON failures on all domains (healthcare, developer tools, finance)
|
| 69 |
+
|
| 70 |
+
## License
|
| 71 |
+
|
| 72 |
+
MIT License - see [LICENSE](LICENSE) file for details
|
| 73 |
+
|
| 74 |
---
|
| 75 |
|
| 76 |
+
**Built with ❤️ for AMD AI Hackathon**
|