Weather API with Ollama Integration

A Spring Boot application that provides weather information using the Ollama API for generating responses. This application demonstrates how to integrate a local AI model (Mistral) with a Spring Boot REST API.

Prerequisites

• Java 23
• Maven
• Ollama installed and running locally
• Mistral model installed in Ollama

Getting Started

1. Install Ollama

Follow the instructions at Ollama's official website to install Ollama on your system.

2. Install the Mistral Model

Run the following command to download the Mistral model:

ollama pull mistral

3. Build the Application

mvn clean install

4. Run the Application

mvn spring-boot:run

The application will start on port 8080.

API Endpoints

1. Current Weather

GET /api/weather/current/{location}

Example: http://localhost:8080/api/weather/current/New York

2. Weather Forecast

GET /api/weather/forecast/{location}?days={days}

Example: http://localhost:8080/api/weather/forecast/London?days=3

Project Structure

• com.arrayindex.weatherlangchain - Main package
  • config - Configuration classes
    • OllamaConfig - Configuration for Ollama API client
  • controller - REST controllers
    • WeatherController - Handles weather API requests
  • model - Data models
    • WeatherResponse - Response model for weather information
  • service - Business logic
    • WeatherService - Service for generating weather information using Ollama

Configuration

The application is configured in application.properties:

server.port=8080
ollama.model=mistral

How It Works

1. The application receives a request for weather information
2. The request is processed by the WeatherController
3. The WeatherService generates a prompt for the Ollama API
4. The Ollama API (using the Mistral model) generates a response
5. The response is returned to the client

Postman Collection

A Postman collection is provided in the Weather-API.postman_collection.json file. Import this file into Postman to easily test the API endpoints.

Notes

• This application uses the Ollama API to generate weather information, not real weather data
• The responses are generated by the AI model based on the prompts
• The application requires Ollama to be running locally on port 11434 "# mcp-test"