Solutions / Deep custom
Proposals that show the projected saving on the client’s own roof.
A system that builds a 3D model of a site, places the panels, and projects generation against consumption hour by hour, so the proposal shows its working, with every assumption stated.
What outside research says
190% is how far South Africa’s national average electricity tariff has risen since 2014, mostly above inflation.
Source: CSIR, 2025 (opens in a new tab)8.76% is the average electricity price increase approved for customers supplied directly by Eskom from 1 April 2026.
Source: Eskom, 2026 (opens in a new tab)
Findings from the research linked under each figure. They are not Yielde results.
The loop
How it works
From a roof and a bill to a signed-off proposal
Seven steps, from the first event to the result. Choose a step to read what happens in plain words, how it is built, and what keeps it right.
Step 1 of 7
A site needs a proposal
In plain words
Your salesperson has a client who wants a proper proposal. They drop in the address, the electricity bills and the notes from the call or the visit.
Safeguard Every proposal is tied to a fixed set of inputs, so any number in it can be reproduced from what was supplied.
Under the hood
It is designed so that each proposal is one project: the address, the bills or meter data, photos, measurements and notes, and the panels and inverters to consider. Once the inputs are confirmed, the modelling is deterministic, which means the same inputs always give the same 3D model and the same numbers. When an input changes later, the new run can be compared line by line with the first.
Works with
- Utility bills and usage data
It reads the bills
In plain words
Electricity bills come in every layout under the sun. It reads them and your site notes, and writes down the monthly use, the tariff and any new loads the client mentioned.
Safeguard A value it could not read from a bill is left empty and flagged for your team. It is never filled in with a typical figure.
Under the hood
The language model reads the bills, often scanned PDFs, and the site notes, and fills a fixed set of fields: monthly use, tariff, demand charges where they apply, roof type and planned loads such as a pool pump or an electric vehicle. Each value links to the page it came from, and values that do not add up are flagged. When only monthly bills exist, the hourly use profile is built from the building type and labelled as an assumption.
Works with
- Utility bills and usage data
It builds the site in 3D
In plain words
It rebuilds the roof and everything around it in 3D: the chimneys and water tanks, the neighbour’s double storey and the big tree on the north side.
Safeguard Where the imagery and your measurement disagree, your measurement is used and the difference is shown to the engineer.
Under the hood
The roof and its surroundings are rebuilt from aerial imagery and height data where it exists, then corrected with your team’s measurements. Each roof face gets an outline, a direction and a pitch. Pitch cannot be read from a single overhead image, so it comes from height data or your measurement, or is listed as an assumption. Chimneys, tanks and skylights are marked as obstructions. Trees and nearby buildings stay in the model because they cast shade.
Works with
- Aerial imagery
It designs and simulates
In plain words
It places panels from your own list where they fit, then works out what they would produce in every hour of a typical year, against how the client uses power.
Safeguard Only panels and inverters from your list are used. No panel goes inside your setbacks (the clear space kept from roof edges) or on a marked obstruction.
Under the hood
Panels and inverters from your list are laid out on the usable faces in a few options. Each option is simulated hour by hour across a typical year: the sun’s position, irradiance (the sunlight reaching each panel) from long-term weather data for the location, the shade the 3D model casts at that hour, heat and wiring losses, and the inverter’s limits. The result is set against the usage profile to show self-use, export and what a battery would add.
Works with
- Your panel and inverter list
It writes the proposal
In plain words
It turns the results into a proposal a homeowner or a finance director can follow, with a page that lists every assumption it made.
Safeguard The language model writes the words, not the numbers. A figure in the text that does not match the simulation stops the draft.
Under the hood
The language model writes the explanation: what the system is, what it is expected to produce through the year, how much of that the building uses itself, what a battery adds, and the payback, worked out by code from the system price, the client’s tariff and the tariff increase you assume. Every number in the text is filled from the simulation and checked before the draft moves on. Production is shown as a range with the main uncertainties named, such as pitch from imagery or a usage profile built from monthly bills.
Your engineer signs off
In plain words
Your engineer looks over the model, the layout and the numbers, fixes what is off and signs it. Only then can the proposal go to the client.
Safeguard No proposal reaches a client without an engineer’s recorded sign-off. The system cannot sign off its own design.
Under the hood
The engineer sees the 3D model, the layout and every input and assumption next to the draft, with flagged bill values first. A corrected measurement or input runs the simulation again, and the earlier version is kept. Export needs their sign-off, recorded with their name and the date. It is an estimate that supports their judgement: it does not replace a site survey or the electrical design and sign-off.
Measurement corrected: goes back to step 03, It builds the site in 3D.
The client gets a proposal
In plain words
The client gets a proposal in your brand that shows its working, and a 3D view of their own roof with the panels on it that they can turn and explore.
Safeguard Every proposal states that it is an estimate made before a site survey. The assumptions page is part of every export and cannot be left out.
Under the hood
The signed-off proposal is exported in your brand as a PDF, with the assumptions and tolerance page included, and a link to the 3D model the client can look around. Inputs, 3D model and results are saved together as one version. When the site survey corrects a measurement, the proposal is run again and the client is shown what changed and why, rather than a new set of numbers with no explanation.
Works with
What it knows
Aerial imagery and any height data for the site, the client’s bills or meter data, your team’s measurements and notes, your panel and inverter list, your design rules and long-term weather data for the location. Each input is stored with its source.
What it may do on its own
Read bills into usage figures, rebuild the site in 3D, lay out panels from your list within your rules, run the yearly simulation and draft the proposal. It can re-run these as often as needed, because none of them reaches the client.
What waits for your yes
The design and the proposal before any client sees them, every input it could not read or that did not add up, and any change after sign-off. The site survey and the electrical design and sign-off stay with your qualified people.
How we test it
Before use it models systems your team has already installed, where the measured layout and metered production are known, and its estimates are compared with them to set the tolerance it states. Every change to the simulation is replayed on those sites before it goes live.
What we measure
Once it is live, these are tracked and reviewed with you.
- Time from site details to a signed-off proposal
- Engineer corrections before sign-off
- Projected generation against metered generation
What you get
- A 3D site model the client can look around
- Hourly generation and consumption projection
- Battery and inverter sizing options
- A proposal document in your brand
Works with
- Aerial imagery
- Utility bills and usage data
- Your panel and inverter list
Who it is for
Solar and energy firms selling larger systems where the numbers close the deal.
Built to order. We scope it with you on a call, price it before any work starts, and run it in draft mode until you trust it.
Related
- Quoting agent
Quotes drafted the same day, priced the way you would price them.
An agent that turns an enquiry, a site photo or a drawing into a priced quote using your own rates, materials and margins, ready for you to check and send.
Built to order - Solar prospecting agent
A ranked shortlist of the roofs to quote first.
An agent loop for solar companies: it reads satellite imagery for roof size and shading, estimates energy use, checks public company records, and hands sales a shortlist of properties that fit.
Built to order - Research and reporting agent
Decisions made on this week’s numbers, not last quarter’s.
An agent that joins your financials to market research and outside data, and writes a weekly brief the way an analyst would: what changed, why, and what to look at.
Built to order