Accurassi

The problem

Home electrification is unstoppable, and selling it is miserable. A household’s options depend on their tariff, their consumption shape, their roof, their orientation, their shading, and what their retailer happens to offer this month. Every one of those is a different data source in a different format.

Accurassi exists so the company selling to that household does not have to solve any of it. They ask a question; we answer it with numbers that hold up.

Solar Seeker

Solar Seeker is the part I build. Given imagery of a property, neural networks segment the roof into usable planes — pitch, azimuth, obstruction, available area — and that geometry feeds a system model that sizes panels and batteries against the household’s actual half-hourly consumption rather than a national average.

The output is a quotable system, not a brochure estimate: what fits on that roof, what it will generate, what it will save, and how long it takes to pay for itself.

Around the model

• Bill extraction and plan optimisation refined since 2018
• Consumer Data Right integration for real, permissioned meter data
• APIs plus a Model Context Protocol interface, so AI agents can consume the same engine
• energymadeeasier.au as a public showcase of the comparison stack