See the difference
Home Assistant exposes raw entities and integrations. Lovelacer reads your registry, infers rooms, and generates a clean dashboard you can preview before applying.
Real screenshots from a Czech Home Assistant install. Your dashboard will use your own rooms, entities, and language.
What you get
Lovelacer uses Home Assistant areas when available, then falls back to room-name detection for entities that were never properly assigned. Every placement includes a confidence score, so you can review uncertain matches before applying.
Re-running analysis shows what changed: new entities, moved entities, and removed entities. Nothing is written to your dashboard until you click Apply.
When Lovelacer can confidently place an entity but Home Assistant still has no room assigned, it suggests fixing the underlying area configuration. Cleaner entities lead to cleaner dashboards over time.
How it works
Lovelacer installs as a Home Assistant add-on and opens directly inside HA. No separate login, no exposed ports, no external service required.
Standalone Docker support is available for Home Assistant Core users.
Lovelacer groups entities by room, scores uncertain matches, and lets you review everything before applying. Rename rooms, change icons, move entities, or hide things you don't want on the dashboard.
Lovelacer creates a separate dashboard using Home Assistant’s native storage mode. Your existing dashboards stay untouched, and nothing is overwritten unless Lovelacer created it originally.
By the numbers
Based on a real Home Assistant install with multiple integrations and years of accumulated entities. Your results depend on how clean your existing setup already is.
The hard parts
Room detection works across English, Czech, German, Spanish, French, Italian, Polish, and Dutch. Lovelacer recognizes localized room names and falls back gracefully when integrations use inconsistent naming.
Every placement includes a confidence score and the signals behind it — Home Assistant areas, device metadata, friendly names, and entity IDs. When something looks wrong, you can immediately see why.
If Lovelacer confidently identifies a room but Home Assistant still has no area assigned, it suggests fixing the underlying configuration instead of repeatedly guessing around it.
Re-running analysis shows additions, removals, and moved entities before anything changes. Lovelacer creates a separate dashboard and never overwrites your existing views.
The base product is fully local and heuristic-driven. Future AI-assisted features remain opt-in and support local Ollama models for users who prefer keeping everything inside their network.
Lovelacer runs as a Home Assistant add-on by default, but the same image also works standalone for Home Assistant Core and Container users.
Under the hood
Lovelacer starts with Home Assistant’s own structure — areas, device metadata, friendly names, and entity IDs — then combines those signals into a confidence score you can review per entity.
Room detection works across English, Czech, German, Spanish, French, Italian, Polish, and Dutch. Localized room names like kuchyně, Wohnzimmer, cocina, or cuisine still land in the right place.
The base product runs entirely locally and never leaves your network. Optional AI-assisted features remain opt-in and support local Ollama models for users who prefer fully local workflows.
From one real install
Same detection logic. Different home, different rooms, different language.
The screenshots across this page come from this install — real entities, real room names, real clutter.
Localized room names resolved automatically without any per-install configuration.
Where we are
Focused on one thing: generating cleaner Home Assistant dashboards from real-world installs.
Bug? Missing language? Confusing behavior? Open an issue — issues from real users shape what we work on next.
Get started
You run HA OS or HA Supervised and have the add-on store. Add the custom repository, click Install, open Lovelacer from the sidebar.
Install guide →You run HA Core, HA Container, or want a local dev runtime. Run the published image with a long-lived access token. Same features.
Install guide →