Mistral 7B LLM Model
In Patronus, the Mistral7B model from Hugging Face powers our AI-based predictions for flight data. This model provides high-performance natural language processing, enabling it to process user flight requests and predict the optimal time for booking based on patterns in historical and live data.
Why Mistral7B?
Mistral7B is a robust language model known for its accuracy in processing complex queries and generating insightful outputs, making it a great fit for Patronus's requirements for precise and efficient flight prediction.
How it Works
Data Transmission: When a user enters flight details (e.g., from and to locations, departure and arrival dates), this data is stored in Appwrite's database.
Model Querying: The flight details are then sent from Appwrite to the Mistral7B model.
Prediction Generation: Mistral7B processes this data and returns a prediction with details on the cheapest flight options, estimated dates for lowest prices, and booking links.
Response Delivery: Once generated, this information is delivered back to the Patronus backend and sent to the user’s mailbox through Appwrite's Mailgun integration, ensuring they receive timely insights.
This integration allows Patronus to leverage advanced machine learning capabilities, providing users with actionable insights for booking flights at the best possible times.
Last updated