"Don't leave money on the table. Let AI handle the deal."
Negotio.ai is an intelligent autonomous agent built for the Autonomous Hack 2026. It solves the "Dead Capital" crisis—where millions of dollars in used goods sit in closets because selling them is too stressful and time-consuming.
Unlike standard chatbots, Negotio.ai is goal-oriented. It doesn't just chat; it actively performs work. It scouts the live web to find real-time market valuations and acts as a ruthless negotiation coach, generating data-backed counter-offers to maximize seller profit.
- Dual-Mode Brain: Instantly switches between "Market Researcher" (finding data) and "Negotiation Coach" (closing deals).
- Live Internet Access: Uses
duckduckgo-searchto find actual 2025 pricing, not outdated training data. - Self-Healing Architecture: Features a "Silent Failover" system that ensures the demo never crashes, even if the search API is rate-limited.
- Assertive Persona: Engineered to be opinionated and decisive, acting like a Senior Consultant rather than a passive assistant.
- Real-Time Valuation: Scrapes eBay and forums to find the true "sold" price of any item.
- Comparison Tables: Automatically generates Markdown tables comparing models, prices, and features.
- Verdict Engine: Doesn't just list options; picks a "Winner" and tells you why.
- Lowball Detection: Instantly flags offers that are under market value.
- Script Generation: Writes professional, firm, and data-backed replies for you to copy-paste to buyers.
- Psychological Edge: Uses price anchoring tactics to protect your profit margins.
- Frontend: Streamlit (Custom CSS for Dark "SaaS" UI)
- Core Intelligence: Google Gemini 2.5 Flash (Low latency, High reasoning)
- Tools:
duckduckgo-search(Live Web Access) - Backend: Pure Python 3.13
- State Management:
st.session_state(Persistent Chat History)
Follow these steps to run Negotio.ai locally.
git clone [https://github.com/yourusername/negotio-ai.git](https://github.com/yourusername/negotio-ai.git)
cd negotio-ai
pip install streamlit google-generativeai duckduckgo-search python-dotenv
Create a .env file in the root directory and add your Google Gemini API key:
GOOGLE_API_KEY=your_api_key_here
streamlit run app.py
User: "What is the best laptop for coding under $800?" Negotio: Scours the web for 2025 reviews, creates a comparison table of the MacBook Air M1 vs. Dell XPS, and gives a final recommendation.
User: "I am selling my iPhone 13. Someone offered me $300."
Negotio: > "
Real Market Value: $450 - $500 Offer Received: $300 (Lowball) Suggested Reply: 'Hi, thanks for the offer. I checked current sales and these are going for $450+. I can meet you at $425, but $300 is too low.'"
A[User Input] --> B{Intent Detection}
B -->|Research| C[DuckDuckGo Search Tool]
B -->|Negotiate| D[Market Price Check]
C --> E[Raw HTML/Snippets]
D --> E
E --> F[Gemini 1.5 Flash (Analysis)]
F --> G[Structured Response (Table/Script)]
G --> H[Streamlit UI]
- WhatsApp Integration: Allow Negotio to reply to buyers directly on your behalf.
- Image Vision: Upload a photo of a garage sale item, and Negotio identifies it and prices it instantly.
- sniper Mode: Automatically find underpriced items on eBay for arbitrage.
Submission for: Autonomous Hacks 2026 Status: MVP Complete (v1.0)
Made with ❤️ by Pair Programmers