The most expensive thing about a frozen warehouse is the moment you forget it exists. Every compressor, every fan, every light hums along, consuming power at whatever price the grid demands that hour. For operators of cold storage and food logistics sites, where cooling can account for up to 90% of the energy bill, this is a relentless, silent tax [Gyre Energy, retrieved 2026]. A London startup called Gyre Energy is betting it can turn that passive cost into an active, optimised asset. Its method is a combination of physics-based AI and a thermal battery, designed to slip into existing facilities and teach the chillers to work smarter, not harder.
Gyre’s proposition is straightforward in a way that appeals to facility managers who speak in kilowatt-hours and payback periods. It installs a software-and-storage layer that doesn’t require ripping out the existing refrigeration plant [The SaaS News, July 2026]. The system’s AI builds a digital twin of the site, learning its thermal dynamics and usage patterns [Gyre Energy, retrieved 2026]. Then, connected to energy market data, it shifts cooling loads to times when electricity is cheapest and greenest, storing that ‘cold’ in its thermal energy storage unit,essentially a large, insulated tank of phase-change material,for use during peak, expensive hours [Gyre Energy, retrieved 2026]. The company claims the setup can cut cooling costs and emissions by up to 40% with a payback period under 18 months [MIT Solve, retrieved 2026] [Greer Clausen - LinkedIn, retrieved 2026].
A wedge into a massive, warming market
The market Gyre is entering is both enormous and heating up. Cooling accounts for 20% of global energy demand, a figure projected to triple by 2050 [Gyre Energy, retrieved 2026]. In the UK alone, cold storage operators spent £1.2 billion on energy last year [Financial Times, 2025]. The startup’s initial wedge is the cold-chain logistics sector, a world of vast warehouses where temperature stability is non-negotiable. Its first reported deployment is with an unnamed global logistics giant, planned for a chamber within a 140,000-square-foot facility [The Next Web, July 2026]. Success here, measured against an IPMVP baseline, is the template for selling into grocery retail, industrial cooling, and potentially data centers.
The Oxford MBA founding trio
The company was founded in 2024 by three Oxford MBAs, a team built around complementary technical and commercial stripes. The backgrounds, as reported, suggest a deliberate focus on the intersection of machine learning and hard energy systems.
| Founder | Role | Reported Background |
|---|---|---|
| Dougald Coulson | CEO | Machine learning and energy technology commercialisation [BusinessGreen, July 2026]. |
| Michael McKenna | CTO | Machine learning and energy systems [IT Brief, July 2026]. |
| Tom Gibson | COO | Energy systems and energy-technology commercialisation [IT Brief, July 2026]. |
They are developing the core technology with academics from the University of Oxford, and the company was a winner of the OxBridge AI Challenge [RWTA, retrieved 2026]. This academic and technical pedigree was likely a key factor for their lead investor.
Why Speedinvest wrote the check
In July 2026, Gyre announced a pre-seed round of more than $1.3 million, led by European early-stage firm Speedinvest with participation from rule30 and Plug and Play [The Next Web, July 2026]. The funding is earmarked to scale from its first pilot into larger enterprise and international cold-chain environments [Gyre Energy, retrieved 2026]. For a hardware-enabled software play, the round is a crucial runway to prove unit economics at a single site before attempting to replicate them. The investor lineup signals confidence in a capital-efficient rollout; the hardware is a storage tank, not a custom chiller, which keeps deployment complexity and cost contained.
The risks of selling to a conservative industry
Selling new technology to industrial operators, however, is rarely a simple matter of superior math. The risks for Gyre are less about the physics and more about the practicalities of adoption in a conservative, risk-averse sector.
- Sales cycle friction. Convincing a logistics company to let an AI control the climate for millions of dollars worth of perishable goods requires immense trust. The sales motion is inherently consultative and proof-driven, which can be slow and expensive to scale.
- The integration puzzle. While Gyre says its platform integrates into existing systems, no two industrial refrigeration setups are identical [Gyre Energy, retrieved 2026]. Each deployment requires custom tuning of the digital twin, a process that must be streamlined to avoid becoming a services-heavy consultancy.
- Established competition. Gyre is not alone in applying AI to building efficiency. A named competitor is BrainBox AI, which has a broader focus on commercial HVAC. Gyre’s differentiation rests on its dedicated focus on high-stakes cold storage and the integration of physical thermal storage, which provides a tangible buffer that pure software cannot.
The company’s most plausible answer to these risks is the promised payback period. An 18-month return is a language every operations director understands. If Gyre can consistently deliver that, the conversation shifts from fear of new technology to a straightforward capital expenditure calculation.
The next twelve months
The immediate plan is clear: prove the model at the first major logistics site. All strategic roads lead from that single 140,000-square-foot chamber. A successful deployment, with verified savings and stability data, becomes the case study to open doors at other cold-chain operators and adjacent verticals like large-scale food retail. The pre-seed funding should carry the team through this crucial proof-of-value phase. A logical next step would be a seed round in 2027 to fund a dedicated sales and deployment team, scaling the playbook across a first cohort of enterprise customers.
The unit economics, in theory, are compelling. Take a mid-sized cold storage facility with a £200,000 annual cooling bill. A 30% saving is £60,000 per year. If Gyre’s system costs £90,000 installed (estimating to hit the sub-18-month payback), the site owner is cash-positive in under a year and a half. That’s a proposition that could get a CFO’s attention even before the sustainability officer walks in the room. The incumbent Gyre must beat isn’t another AI startup; it’s the inertia of the existing, inefficient, always-on chiller. Their job is to make that chiller look like a liability that’s quietly burning money, one expensive kilowatt-hour at a time.
Sources
- [Gyre Energy, retrieved 2026] Company Website | https://www.gyreenergy.com/
- [The Next Web, July 2026] Gyre Energy raises $1.3m to make the world’s cold stores work less when power costs most | https://thenextweb.com/news/gyre-energy-cooling-funding
- [BusinessGreen, July 2026] Gyre Energy raises $1.3m to scale AI-powered industrial cooling technology | https://www.businessgreen.com/news/4532406/gyre-energy-raises-usd13m-scale-ai-powered-industrial-cooling-technology
- [IT Brief, July 2026] Gyre Energy wins funding for larger cold chain rollout | https://itbrief.news/story/gyre-energy-wins-funding-for-larger-cold-chain-rollout
- [The SaaS News, July 2026] Gyre Energy Raises $1.3M Pre-seed | https://www.thesaasnews.com/news/gyre-energy-raises-1-3m-pre-seed/
- [MIT Solve, retrieved 2026] MIT Solve Profile | https://solve.mit.edu/
- [Greer Clausen - LinkedIn, retrieved 2026] LinkedIn Post on Performance Metrics
- [Financial Times, 2025] Article on UK Cold Storage Energy Spend
- [RWTA, retrieved 2026] Supplier Profile | https://www.rwta.com.au/find-a-supplier-source/gyre-energy
- [Dealroom, July 2026] Gyre Energy raises $1.3M pre-seed to cut industrial cooling bills with AI | https://dealroom.co/news/137289-gyre-energy-raises-1-3m-pre-seed-to-cut-industrial-cooling-bills-with-ai/