A smart, context-aware travel companion that prepares you for your journey using Real-Time Data and Local AI.
This project was developed at The Paisii Hilendarski University of Plovdiv for the Software Quality Assurance (Q.A.) course.
Unlike standard cloud-based apps, HotelMind is designed to run with Local AI to ensure privacy and customizability.
- Supported Engines: Ollama or LM Studio.
- Model Requirements: The selected AI model must support Tool Calling / Function Calling (e.g.,
mistral-nemo,qwen3.5:4b, orllama3-groq-tool-use) to interact with the search and weather APIs. - Endpoint: Ensure your local server is running (usually
localhost:11434for Ollama) so the Semantic Kernel can orchestrate the requests.
Find stays within your specific budget. The AI searches the web in real-time to find current prices and availability, summarizing the best options for you.
Get a live 7-day forecast for your destination. It doesn't just show numbers; it uses current conditions to power the rest of the app's intelligence.
The AI analyzes the High and Low temperatures of your destination to suggest the perfect gear.
Get an AI-generated "feel" for the city. Discover the local atmosphere and top-rated activities summarized in a clean, readable widget.
| Layer | Technology |
|---|---|
| Backend | ASP.NET Core MVC (C#) |
| AI Orchestration | Microsoft Semantic Kernel |
| Local AI | Ollama / LM Studio |
| Search API | Serper.dev (Google Search) |
| Weather | WeatherAPI.com |
| Frontend | CSS3 Glass-morphism, JavaScript, Bootstrap 5 |
