How Igniwave works.
A step-by-step look at the current build, from first setup to asking the local model useful questions about your week.
Download your private AI
On first launch, you create the single local workspace account for this browser. Nothing is provisioned in the cloud.
Download your local AI model
Choose a model and download it to your device. The initial download can take a few minutes, but afterward inference stays local.
Add simple calendar events
Add plans, appointments, routines, or personal blocks in the calendar. These events become part of the shared timeline context.
Connect Oura or manual data
Device connections store non-secret provider state locally and normalize the metrics into one format the rest of the app can read.
Save daily check-ins and notes
Morning or evening check-ins capture mood, energy, stress, focus, and sleep quality. Quick notes land beside them in the timeline.
Ask the local AI about your week
Ask for a prep brief before an event, look for patterns across the last week, or ask what changed. The answer comes from your timeline, not a generic prompt.
What the AI can and can't do
What it does: The AI reads your calendar context, check-ins, notes, metrics, and saved memories to answer grounded questions about your own data.
What it doesn't do: It does not send your prompts to a cloud model, it does not invent missing facts, and it does not replace professional advice. If context is missing, it should say so.
System requirements
Igniwave requires a browser with WebGPU support for local AI inference:
- Google Chrome (desktop or Android) — recommended
- Microsoft Edge (desktop)
- Windows, macOS, or Android
Safari on iPhone and iPad does not yet support WebGPU and is not currently supported. Firefox is not supported.