Impact
AI is becoming one of the fastest-growing sources of energy consumption on the planet. Every question asked of a model, every document summarized, every line of code generated burns real electricity in real data centers. And the scale is exploding: large companies now run millions of these requests a day, and nobody inside those companies can see what's actually being used, by whom, or why.
The problem isn't just cost. It's waste. Research from the AI Energy Score project — which our CEO Richard co-founded with Hugging Face — has shown that some AI models burn up to 6,000× more energy than alternatives that would do the same job just as well. Companies are paying for that excess, and the environment is absorbing it, because the waste is invisible. You can't manage what you can't measure.
Argmin is building the measurement layer. We deploy inside a company's own systems and trace every AI request back to the team, service, and workload that triggered it. Once the waste is visible, it can be cut. Our early analysis suggests 15–25% of enterprise AI spending is going to models that are overpowered for the job they're doing. Eliminating that waste means lower bills, lower energy use, and lower emissions — at the scale of the largest AI consumers in the world.
The impact compounds. Every company we help becomes a template for the next. Every model swap we enable reduces demand on the data centers driving AI's footprint. And the same measurement system we're building for cost today is designed to support full carbon accounting and sustainability reporting next — giving companies, regulators, and the public a real answer to the question "what is AI actually costing the planet?"
We're two founders with deep backgrounds in enterprise infrastructure and climate research, building the layer that makes AI consumption accountable.
What I'll do with $5,000
The product is built. What $5,000 buys us is the infrastructure to stand it up in a live, demo-ready environment — AWS hosting, domains, the deployment stack — so we can put it in front of our first design partners without asking them to take a leap of faith. We're not asking for funding to invent something. We're asking for the runway to show it working.
Quick Bio
Graduated bachelor's as valedictorian at 13. Now an Oxford PhD in AI. Met Richard there with a nagging feeling that AI's environmental cost was invisible — and nobody was going to fix it but us.