Generate New Forecast

Forecast Configuration
Select the electricity market area for the forecast.
Choose the AI model for forecast generation. OpenAI requires API key and internet connection. Ollama uses local AI model via Ollama service. Hugging Face uses a local transformer model (requires model download, runs on CPU/GPU).
Warning: This will generate a forecast using simulated data instead of real market prices. Use this option only for testing when real data is unavailable. Results will NOT be accurate for actual trading decisions.
Forecast Process:
  1. Scrape current market analysis from Tibber
  2. Fetch historical spot prices for the past week
  3. Generate AI-powered 7-day hourly forecast using selected AI model
  4. Store results in database for analysis
Back to Dashboard
How It Works
Data Collection

The system automatically scrapes the latest market analysis from Tibber's website and fetches historical spot prices from Nord Pool. Real market data is required for accurate forecasting.

AI Analysis

The selected AI model (OpenAI GPT-4, Ollama, or Hugging Face) analyzes the market conditions and historical patterns to generate a detailed 168-hour (7-day) forecast with hourly price predictions.

Data Storage

All forecasts and historical data are stored in a structured database, enabling trend analysis and forecast accuracy evaluation over time.

Visualization

The results are presented in an interactive dashboard with charts, statistics, and API access for integration with other systems.