Janela — a local Gemma tells you when to go outside (and why)

Description and use

Janela offers precise outdoor activity windows by scoring weather data in Java and generating friendly explanations with Gemma.

Problem

People need a clear recommendation of when to go outside for specific activities, but weather apps only provide raw data.

Outcome

Janela delivers a deterministic, data‑driven recommendation for outdoor activities, combining Java scoring with a small LLM for natural language explanations.

Workflow

• 1. User selects city, activity, duration, and look‑ahead days.
• 2. Spring Boot API calls Open‑Meteo for geocoding and hourly forecast.
• 3. WindowScorer (Java) evaluates each daylight hour, scoring for rain, temperature, UV, and wind.
• 4. Top three windows per day are ranked; the best overall window is chosen.
• 5. If Ollama is available, Gemma writes a friendly explanation; otherwise a template is used.
• 6. Results are returned with formatted day names, times, and rounded metrics.
• 7. User receives a shareable link to the recommendation.

Evidence

Janela does that for you. You give it a city, an activity (run, walk, bike, picnic, gardening, or an outdoor workout at the square's gym), how long you want to be out, and how many days ahead to look.

Source

city + activity ──▶ Spring Boot API ──▶ Open‑Meteo (geocoding + hourly forecast) │ ▼ WindowScorer (plain Java, unit‑tested) │ top 3 windows, already ranked ▼ Ollama up? ── yes ──▶ Gemma writes the explanation └── no ───▶ template writes it instead

Source

The model never picks a time and never does math.

Source