AI-native quoting platform

For UK solar & battery installers

AI sells the system. Half-hourly dispatch simulation proves the numbers.

An AI chatbot answers every customer question. A lead-scoring engine tells you which deals to chase. An MCS auditor blocks non-compliant sends. A Proposal Builder learns your style from a single PDF. All on top of the most rigorous UK battery + PV dispatch simulator there is— 17,520 simulated half-hour slots per year, against real smart-tariff data and real PVGIS irradiance, rolled forward 25 years with battery fade + panel degradation + tariff inflation.

Try a quote →Settings

17,520

Half-hourly dispatch decisions per simulated year

25 yr

Roll-forward with battery fade + panel degradation + tariff inflation

11+

AI features live or shipping

AI on every front

Not a feature we bolted on · how the platform thinks

AI is woven through every step — intake, customer chat, PDF parsing, MCS audit, lead scoring, follow-up emails. Built on Claude (Anthropic’s top model), fast enough for the customer to chat live, included at every paid tier.

Most-used AI feature

Customer chatbot, trained on every quote you send

The chatbot has the full quote in context — your customer can ask anything from “why this battery size?” to “what happens if I get an EV in 3 years?” and get a real answer in seconds, in plain English, branded as you. Saves you 4–8 phone calls per quote.

QWill I still save money if rates fall?

AAt 15% rate fall your saving drops from £1,420 to £1,180/yr — still 8.2 yr payback.

QWhat if I add an EV later?

AThe 5 kWh battery covers ~70% of EV overnight charging on a Go-style tariff. Recommended.

AI Inside

Lead scoring engine

AI analyses chat transcript, return visits, scroll depth, urgency words, commitment language → conversion score 1.0–10.0, recalibrated against YOUR signed-vs-lost history.

7.8 / 10“when can you install” detected

Claude vision

Proposal Builder learns YOUR style

Upload your existing PDF, Claude vision extracts palette, typography, section structure. Chatbot iterates with live preview (“use our blue”, “move payment to page 2”). AI suggests improvements based on what converts.

Blocks bad sends

MCS compliance auditor

16 checks before every send — Performance Estimate maths, sun-path shading, Product Directory cross-reference, EESS class, signature block. AI parses your Reonic PDF, validates against MCS rules, blocks non-compliant sends.

Claude analysis

Conversation intelligence

Per-turn sentiment, extracted topics, buying signals (urgency, commitment words), objections detected. You see at a glance whether the customer’s warming up or going cold.

Re-engagement

AI-drafted follow-ups

Customer viewed but didn’t accept? AI drafts a personalised re-engagement email referencing the specific package they spent time on. One-click send. Tracks if it converted.

Also · ambient AI doing the grunt work

Intake chatbot

Gathers customer details over a chat, not a form

Bill image OCR

Drop a utility bill PDF → usage extracted in 5 s

PDF cross-check

Reonic PDF parsed, every figure cross-checked against the estimate

SOW interview

Engineer chatbot walks through discrepancies, generates the SOW

Our moat

The most accurate battery modelling in UK quoting

17,520 dispatch decisions / year · 25-year roll-forward

Most quoting tools assume a flat 28 p/kWh and call it a day. That misses the entire point of a battery: the savings come from when you charge and when you discharge, against tariffs whose price changes every 30 minutes. We model it the way it actually works.

A typical winter weekday on a 9.5 kWh battery + smart tariff

One of 17,520 half-hours we simulate

Solar generationBattery chargeBattery dischargeGrid importGrid export

17,520

Half-hourly dispatch decisions per simulated year

25 yr

Lifetime roll-forward with fade, inflation, degradation

Smart-tariff scenarios swept per recommendation

Battery controllers matched to dispatch capability

What flat-rate tools miss

The cheap overnight window is where the savings actually live — flat-rate maths can't see it.
Battery degrades ~2 %/yr; flat models hide it and inflate lifetime savings 25-40 %.
Panel output drops 0.5 %/yr; ignored, the 25-yr cash flow is fiction.
Tariff inflation over 25 years dwarfs the modelled saving — must roll forward year-by-year.
Some battery controllers can do time-of-use arbitrage; some can't. Wrong pairing = wrong savings.
Oversize batteries (2×, 3×, 4× vs base) pay back fast on export-tariff arbitrage — invisible to flat models.

What we model

Real half-hourly rates fetched live, postcode-specific, for every smart-tariff variant.
Battery fade applied per manufacturer's actual cycle curve — Sungrow, Fox, GivEnergy, Tesla, etc.
Panel degradation curve from the module's datasheet (typically 0.45-0.55 %/yr).
Year-by-year tariff inflation built in — cap-history-aware, not a flat % guess.
Per-controller dispatch — Predbat / Sigenergy AI / Tesla Energy Plan / Fox manual / GivEnergy native each modelled separately.
Recommender sweeps battery size by family (1×, 2×, 3×, 4×) and reports payback for each — installer sees the sweet spot.

What’s real vs what’s modelled

Real, live half-hourly

Live smart-tariff rates (via supplier APIs) and PVGIS solar irradiance for the customer’s postcode.

Modelled, not measured

The customer’s half-hourly consumption profile is synthesised from their annual kWh + occupancy patterns — we don’t pull smart-meter data (yet, see roadmap).

Simulated

The dispatch engine itself — 17,520 decisions/year deciding when to charge/discharge against tariff + load + generation, per the customer’s actual battery controller capabilities.

On the roadmap:smart-meter consumption API integration (via n3rgy and similar providers) so we use the customer’s measured half-hourly usage instead of a modelled profile — closes the last assumption in the chain.

Battery cycle fade

Per-manufacturer cycle curves applied year-by-year. Sungrow, Fox, GivEnergy, Tesla, etc. — each modelled with its own real degradation profile, not a generic 2%/yr.

Panel degradation

Module-specific degradation curve from the datasheet (typically 0.45–0.55%/yr). Year 25 output realistic, not a fantasy.

Tariff inflation

Roll-forward year-by-year using Ofgem cap history + a configurable inflation rate. The £ figure in year 10 isn't the same as year 1.

Controller-aware dispatch

Catalogue carries gridChargeMode, gridExportMode, agileOptimisation, setupSummary per controller. Predbat customers get Predbat. Tesla Energy Plan customers get Tesla Energy Plan. We don't pretend they're the same.

Oversize-battery ROI

The 1× recommendation is just one row. The sweep also reports 2×, 3×, 4× battery configs with their own paybacks — for customers who'd benefit from arbitrage on Go + Outgoing Agile.

Defensible numbers

Every £ figure is derivable from real data — PVGIS irradiance, live tariff rates, supplier wholesale prices, your markup. Items without wholesale pricing get an “indicative” tag. Surfaceable to the customer in plain English.

What we do

We sit between your design tool (Reonic, OpenSolar, Easy PV) and your customer. You configure a system, our recommender sweeps every viable combination of inverter, battery, modules, tariff and dispatch strategy, runs a year-long half-hourly dispatch simulation against live smart-tariff data, and produces a branded customer share view your customer reviews and accepts in their own time. Once they’ve picked, you turn it into a formal quote through our cross-checked, chatbot-guided workflow that plays nice with your existing design tool.

Where it hurts — and how we fix it

Six pain points we hear over and over from installers, and the specific things our app does about each.

The pain

Quoting eats half a day per customer.

Site visit, spreadsheet, propose, tweak when they ask about a bigger battery, re-propose.

What we do

Configure once; the recommender sweeps every combination.

Inverter × battery × module count × tariff × strategy, all evaluated in ~5 seconds against half-hourly dispatch and live tariff data. Pick 1-4 packages, send.

The pain

Customers don't understand the numbers.

They get the headline saving but glaze over at the methodology — and then ghost you for two weeks.

What we do

Branded customer view explains the saving in plain English.

Where the money comes from, how the lifetime maths is calculated, a chatbot that answers their specific questions ("Will I really save £1,420 a year?"). You see when they viewed it, what they picked, when they accepted.

The pain

Real tariff economics are hard to model.

Most quoting tools use a flat 28 p/kWh assumption that misses where battery savings actually live — the cheap overnight window and the evening peak. They also hide 25-40% of the headline saving by skipping battery fade and panel degradation.

What we do

Half-hourly dispatch over 25 years. Real fade. Real inflation.

17,520 dispatch decisions per simulated year, year-by-year roll-forward with per-manufacturer battery fade + panel degradation curve + tariff inflation. Per-controller dispatch (Predbat ≠ Tesla Energy Plan). Numbers your customer can trust.

The pain

Maintaining two systems is a tax.

You quote in one tool, build in Reonic, store the customer in your CRM, email the PDF, chase the signature.

What we do

Two-way integration with the design tool you already use.

Reonic is fully wired (read project on intake, push our recommendation as a pinned note, parse the engineering PDF, cross-check on the way back). OpenSolar + Easy PV next.

The pain

Compliance work is repetitive.

Pull the MCS Performance Estimate, sunpath shading, BOM, T&Cs — every quote, every time.

What we do

We pull everything out of the Reonic PDF and cross-check it.

Engineer chatbot walks the engineer through any discrepancies + a 10-item Scope of Works interview with yes/no chips. Final quote assembles everything, branded in your colours.

The pain

You can't tell what the customer is thinking.

Sent the quote; the customer's gone quiet; you don't know if they read it, what they picked, or whether to chase.

What we do

Activity log + acceptance trigger.

See when they opened it, which package they explored, when they accepted. Acceptance fires a high-priority notification that takes you straight to the formal-quote workspace.

Plays nice with your design tool

We don’t replace your engineering design tool — we wrap it. You keep your MCS-compliant design environment; we handle the customer-facing layer.

Reonic

Shipped

Full v3 API: read residential projects, write notes + projects + tasks, push to activity feed, two-way contact sync coming.

API docs →

OpenSolar

Next-up

Project import, components push, customer-action webhooks, optional embedded 3D design via the SDK.

API docs →

Easy PV

Next-up

UK-MCS-focused tool with MIS 3002 structural + MCS 020 shading baked in. Heatpunk shares the same API for heat-pump installers.

API docs →

How it works

1
Initial estimate
Configure or import the customer. The recommender ranks every viable system combination against a year-long half-hourly dispatch sim. Send 1-4 packages as a branded share link.
2
Customer reviews
They open the link, compare options, ask the AI chatbot questions about their specific numbers. You see every view + every selection. When they're ready, they tap Accept.
3
Formal quote
Acceptance routes you to the workspace. Push the components to Reonic. Engineer builds it there, generates the proposal PDF. Upload back to us. Claude parses it, cross-checks every figure, the chatbot walks you through any discrepancies + the Scope of Works. Generates the branded final quote for the customer to sign.

What your customer sees

Branded as you · live today on every quote

When you hit send, your customer gets a private share link to a full interactive portal — not a PDF buried in their inbox. They explore the numbers in their own time, ask the chatbot whatever they want, compare packages, and accept when ready. You see every interaction live.

energy-modeller.app/q/a1b2c3

Sunrise Solar · for Rob Duncan

Branded as you

Annual saving

£1,420

Payback

9.4 yr

Lifetime

£28,500

Pick your package

ValueRecommendedPremium

Ask anything

Will I really save £1,420 a year?

Based on your 4,200 kWh annual use and the live half-hourly tariff for your postcode — yes. The 9.5 kWh battery covers ~73% of your evening peak from cheap overnight rates, and the panels offset your daytime base load.

Your brand, your domain feel

Logo, primary + secondary colours, company name everywhere. White-label option drops our footer entirely (Studio + Enterprise).

AI chatbot, answers in their words

Customer asks anything — tariff choice, battery sizing, payback maths, install timing. Claude answers using their specific quote, with prompt caching keeping cost negligible.

Compare up to 4 packages

Value / Standard / Premium / your-call. Customer sees side-by-side savings, sliders for tariff scenarios. They pick when they're ready.

You see everything live

When they opened it. Which package they spent time on. What they asked the chatbot. If they accepted. Notifications hit your phone — "Rob Duncan accepted Premium · 11:42pm."

Behind the scenes — what you see as the installer

Just now

Rob viewed Premium · 4th time

11 min ago

Asked chatbot about export tariff

2 hr ago

Filled questionnaire · 5/5 answers

Yesterday

First view · 3 min on Standard

Coming very soon

Active development · shipping over the next 6–10 weeks

StandardPremium adds AI

Proposal Builder

The proposal you 100% want, every time, and 100% MCS-compliant. Upload your existing preferred PDF; AI extracts the visual style; chatbot iterates (“use our blue”, “drop the climate page”, “move payment schedule to the front”) with live preview.

MCS auditor · 16 checks

MCS Performance Estimate
Sun-path shading
Product Directory cross-ref
EESS Class 1–4
+ 12 more — blocks any failing send

Premium conversion engine

· AI suggestions on customer behaviour
· A/B testing + funnel analytics
· Industry benchmarks
· Auto follow-ups for non-converters
· Embedded BNPL / green-loan options

Reonic on autopilot

Customer accepts → Reonic project auto-created, tagged, assigned to a real user, components handed off, site-survey + DNO + ordering tasks pre-populated. Polls for the engineering PDF the moment your engineer finishes the build.

Plays withReonic

Premium

Lead scoring engine

Every project gets a 1.0–10.0 conversion-likelihood score, calibrated against your own signed-vs-lost history. Tells you which 5 of your 30 open projects to chase today.

Rob Duncan · KY7 6GQ

7.8 / 10

+2.1 accepted Premium · +1.4 “when can you install” · +1.2 returned 4×

Premium

Online signature + instant deposit

Customer signs the formal quote in-app (canvas-draw, full audit metadata: IP, timestamp, T&Cs hash, exact figures hash — what they agreed to is reproducible forever) and pays their deposit immediately via Stripe. Funds settle to your connected account. Cash flow accelerated by 30–90 days vs invoice-and-wait. Per-tech deposit stages configurable (PV 25/35/40, heat pump 30/30/30/10 with MCS flow-test milestone, EV 100% on completion).

Powered bystripeAudit-grade signature

Premium

Books on auto-sync

Connect once. Signed quote auto-creates an invoice with the right per-tech VAT rates (0% UK residential PV/battery/HP through March 2027, 20% EV). Stripe payment auto-marks invoices paid. End of double-entry admin.

ConnectsQuickBooksXero

Pricing

Per project sent · one share-link delivered = one project (revisions don’t double-count)

Free

Up to 10 leads / mo

Standard

£0Permanent free tier

Premium

Not available on this tier

Half-hourly dispatch sim, branded share view (with our footer), 1 user.

Solo

Up to 75 leads / mo

Standard

£79/mo

Premium

£149/mo

Everything in Free + AI chatbot, multi-package send, formal-quote workflow, 1 design-tool integration.

Try it on a real customer

The fastest way to evaluate the platform is to build a quote for someone you’re currently talking to. ~10 minutes from form to share link.

Start an estimate →Set up your brand & integrations

AI on every front

Customer chatbot, trained on every quote you send

Lead scoring engine

Proposal Builder learns YOUR style

MCS compliance auditor

Conversation intelligence

AI-drafted follow-ups

The most accurate battery modelling in UK quoting

What we do

Where it hurts — and how we fix it

Plays nice with your design tool

Reonic

OpenSolar

Easy PV

How it works

Initial estimate

Customer reviews

Formal quote

What your customer sees

Coming very soon

Proposal Builder

Reonic on autopilot

Lead scoring engine

Online signature + instant deposit

Books on auto-sync

Pricing

Try it on a real customer