Quiet Hours is the household-altitude half of the GridPath / Quiet Hours pair: a mobile-first energy timing product that uses location, weather, grid, rate, emissions, and household-load context to explain why tonight matters and what to shift. It brings the GridPath question down to household altitude, with language, pacing, caveats, and recommendations a person could actually use.
How the home story explains why tonight matters before asking for action.
How demand, renewable drop, grid stress, and household context shape the recommendation.
How the Trust surface keeps source links, privacy posture, timestamps, and limits visible.
How provenance language distinguishes live, delayed, fallback, modeled, and collecting states.
Quiet Hours frames the evening as a sequence: solar fades, routines start, demand rises, price and emissions shift. The interface makes timing feel concrete instead of asking people to decode grid signals. That matters because behavior change rarely starts with a number. It starts when the product can explain why this moment matters in language that does not make the user feel behind.
The tap-the-city interaction makes the advice local. Location setup changes what the product can know, explain, and recommend. Quiet Hours reads from GridPath's operator-facing API, then layers in balancing-authority data, weather, emissions factors, utility-rate structures, modeled household-load defaults, Census context, and permission-based or manual location. The product has to say what it knows and what it is inferring, or the recommendation becomes a beautiful fake line.
The household-load defaults are modeled, not measured, and that distinction matters. The design lesson is bigger than energy timing: recommendations should carry their source health with them. If a value is live, delayed, modeled, fallback, or still collecting, the product should say so. I would rather make a caveat visible than let a polished sentence overstate what the system knows.