How green is my home today?
A 0–100 Green Home Score from seven fixed-weight components, with every point explained.
Lloyds Green Home Score & Retrofit Engine
Know how green your home is, and what to do next.
Trained models calculate every number. The AI only explains them, and every number it writes is checked.
S = Σ wₖsₖ over seven componentsWeights sum to 1
What should I improve?
Only retrofits that suit this home, ranked High/Medium/Low and ordered so insulation comes before a heat pump.
How much could I save?
Electricity and gas worked out separately, with interactions between upgrades, then carbon from DESNZ factors.
Could it add value?
An indicative value range from five hedonic models trained on Land Registry sales, never a single number.
Models calculate. The LLM explains.
Models and calculations
Gradient-boosted models trained on 30M EPC certificates, 50k metered NEED homes and 3.2M Land Registry sales, plus a sequential per-fuel savings simulator and DESNZ 2026 emission factors.
Grounded explanations
The AI sees only the computed report. Every number in its answer is matched against that report. If one is unsupported it gets one retry, and after that the deterministic template is used.
What-if = full re-run
Ask “what if I only add solar?” the AI picks the scenario from a fixed list, the engines recompute everything, and only then are the new numbers explained.
From open data to your dashboard
- Property, energy and sales data
- Feature engineering
- Green Home Score
- Efficiency detector
- Retrofit engine
- Savings, carbon and value models
- AI explanation
- Customer dashboard