The UK has the technology to retrofit its homes, and pension funds are keen to pay for it. What it lacks is a way to connect that money to millions of individual houses. Greenflip, a Carbon13 portfolio company led by a founder who spent five years in council housing, prices the retrofit return on any UK home in about ten seconds, without a survey, turning thousands of homes into something lenders can underwrite. It already has paying council customers.
In 2024, 44% of homes in England were still rated below EPC band C, according to the English Housing Survey. Amandeep Kalra knows that number from the inside. As a council officer, he was handed an estate of 29,000 homes to decarbonise, spread across ordinary streets and several types of tenure, with nowhere near the cash to do it.
The pension funds he spoke to loved the idea. Retrofitted social housing was exactly what they wanted to own: ethical, predictable and inflation-linked. Nobody had a way to get their money into thousands of homes that each needed something slightly different. Greenflip is his answer, and on the latest episode of Angel Insights I asked him to make the case for it.
Retrofit hasn't scaled because the person who pays for it usually isn't the person who benefits. A landlord funds the work, and the tenant gets the lower bills. Grants have covered pilots of 100 or 200 homes, but beyond that a landlord needs a return of their own, and a reliable figure for what it is.
The money is there. The National Wealth Fund has put £1.3 billion of guarantees behind social housing retrofit loans from NatWest, Lloyds, Barclays and THFC, unlocking up to £1.65 billion of lending. Amandeep's read is that it reaches the largest landlords first, the ones with treasury teams and budgets for consultants to build the case. Everyone else is stuck at the pilot.
So the capital exists and so do the homes. What's missing is the data that connects them, at a cost that works for thousands of properties rather than a hundred. That's the gap Greenflip was built to fill.
At pre-seed, the team is most of what you're backing, and Amandeep has spent his career on the buyer's side of this problem.
He started young, selling Kenwood blenders on his dad's market stalls at 11 or 12 (Emma Maslen's six sales checks angels make explain why that matters). He trained as an architect, then moved into the public sector, designing affordable housing in East London and delivering 2,000 homes on a £1 billion development programme before that 29,000-home estate landed on his desk.
Today he also sits as a non-executive director of a housing association, the kind of organisation Greenflip sells to. When he wanted to know exactly how EPC data is collected, and where it goes wrong, he and his co-founder trained as EPC assessors themselves.
That co-founder, Thomas Quiroga, is a software engineer who had already been through the problem as a customer. His own house was quoted at £11,000 a year to run; after retrofitting it, Amandeep told me, this should be Thomas's first year with no energy bill at all. They met on day one of the Carbon13 venture builder.
Portfolio analysis usually takes one house type, works out the retrofit case once and applies it everywhere. Amandeep's problem with that:
So Greenflip scores every home individually. It models two things: how much energy the home uses, and how much it could generate. It then weighs the cost of an upgrade (solar panels, a heat pump, and a battery) against the savings and the uplift in value, producing a return on investment for each address in about ten seconds.
The consumption side combines EPC, Ordnance Survey and Land Registry data with real energy data from thousands of homes tagged to exact addresses. The generation side uses a computer vision model trained on thousands of aerial images that reads a roof the way a solar installer would and automatically places panels. Amandeep says it outperformed Google's solar API within three months, because Google built for the whole world and Greenflip is going country by country.
Greenflip targets 80% accuracy, not 100%, which would need a surveyor and a damp sensor in every room. Customers have told the team 80% is enough for investment-grade decisions across thousands of homes.
The model's real-world energy data is the start of a data moat, and the team is working to deepen it. Clients kept asking whether Greenflip could install the upgrades too, so it is testing partnerships with installers and monitoring how homes perform afterwards. Each installed home sets the prediction against the result, and feeds the gap back into the model.
Greenflip sells to local authorities, housing associations and institutional investors, on a subscription priced by the number of homes they analyse each year. These are slow, sceptical buyers, constantly pitched by startups, and the usual question is whether a two-founder company can get through the door.
Greenflip has. One council went from first call to an invoice paid in Greenflip's account in two months. When Amandeep worked in local government, it could take two months just to raise a purchase order. Others have taken around eight months, which is closer to the norm.
Those installation requests are a signal in themselves: clients who have seen the numbers want Greenflip to deliver them.
Customer timelines and model performance are as Greenflip reports.
Greenflip's job today is underwriting: putting a credible return on each home so lenders and investors can price the risk. The larger prize is helping to fund it.
Amandeep puts the potential return on retrofit as a financial asset at between 6% and 12% a year: local pension funds tackling fuel poverty in their own area might accept the lower end, larger international investors would want the top, and both sit where infrastructure funds operate. The catch is scale. A hundred projects is too small for that capital, so homes need aggregating into packages an institution will buy, which is exactly what portfolio-wide data makes possible.
Further out, he wants Greenflip to pay the upfront cost itself and offer retrofit as a service. He was clear this is still being explored: "Obviously, we're not there." The open question for Greenflip is whether it can move from pricing retrofit to financing it, a step that needs far more capital than selling software.
Asked about the future of energy, Amandeep's answer was decentralised: homes as small power stations, generating, storing and sharing energy with each other and the grid. His comparison is striking. One small modular reactor might generate around 3.7 terawatt-hours a year. By his estimate, you'd need to retrofit about 700,000 homes to produce a similar amount.
The homes, he argues, also cut fuel poverty, reduce spending on grid upgrades, ease NHS costs linked to damp and mould, and make for a more resilient system. I'll admit the comparison flatters rooftops on one count: a reactor runs through a still January night and a solar panel doesn't. Amandeep doesn't claim otherwise. He wants the grid kept for what can't supply itself (hospitals, heavy industry, data centres) while homes with batteries act as a virtual power plant.
Amandeep didn't arrive at Carbon13 with a company. He arrived with a problem, and met Thomas on day one. That's the Carbon13 model: build teams around big climate problems, then back the few that work.
The funnel is steep. Around 600 people apply, about 80 get a place, and 6 to 8 teams take investment, so roughly one in a hundred applicants ends up in a funded company. Since 2021, Carbon13 has backed 107 companies, with around 65% going on to raise follow-on funding across its first six to seven funds. Past performance is not a reliable indicator of future results.
Greenflip is one of them. Carbon13 SEIS Fund X doesn't hold it: Fund X backs the next teams to come through the same process.
Why hasn't home retrofit scaled in the UK? The money exists, including £1.3bn of National Wealth Fund guarantees for retrofit lending. But landlords pay while tenants save, and housing stock is too fragmented for capital to reach individual homes. In 2024, 44% of homes in England were still below EPC band C.
What does Greenflip do? Greenflip, a Carbon13 portfolio company, is an underwriting engine for retrofit. It combines public datasets, aerial computer vision and address-level energy data to score the retrofit return on any UK home in about ten seconds, without a survey, targeting 80% accuracy.
Who buys it? Councils, housing associations and institutional investors, on a subscription priced by the number of homes analysed. One council went from first call to paid invoice in two months, against around eight months at others.
What's next for Greenflip? CEO Amandeep Kalra wants to move from pricing retrofit to financing it, paying the upfront cost and offering retrofit as a service. He's clear it's still being explored.
Carbon13 SEIS Fund X invests in 4 to 10 pre-seed climate-tech companies built through the Carbon13 venture builder, from a £10,000 minimum. Fees apply. SEIS offers 50% income tax relief on up to £200,000 a year, provided you claim it, and tax treatment depends on your individual circumstances and may change.
The current round closes on 6 November 2026. Find out more about Carbon13 SEIS Fund X or read our guide to SEIS investing.
Please note: our office hours are weekdays, 9.30am - 5.30pm.
