- polen859
- Aug 27
- 9 min read
When AI Meets Youth Climate Action

With multi-agent AI and a bias for rigour, what started as a student initiative, Carbon2Capital, translates messy bills and invoices into clear carbon signals and practical steps, then connects users to high-quality climate ventures.

We had the chance to sit down with the founding team of Carbon2Capital, a group of young entrepreneurs using AI to make climate action more personal and practical. What began as a hackathon idea has grown into a platform that translates Scope 1–3 emissions into relatable insights, connects everyday choices with systemic change, and links users to verified green investments. Along the way, they’ve tested their vision through incubators, bank partnerships, and the Hult Prize, learning how to balance innovation with responsibility. As part of our Green Job Chronicles series, their story demonstrates how the next generation of climate tech founders is reshaping both the tools we use and the values on which we build them.
You started Carbon2Capital as university students. What drew you to AI for climate action, and when did tech start feeling like a tool for systems change?
For us, early on, we started realising the impact tech, especially AI, can have on problems we face. We started participating in hackathons, which were always centred around a problem that exists in society, whether that was food waste, health, labour, or sustainability. We had to create a product that solved that problem, so tech was usually beneficial there. For climate action specifically, we began exploring tech in climate action after participating in a hackathon where we were encouraged to utilise AI for a more sustainable future.
What strengths do each of you bring from tech to strategy to storytelling, and how do you work together to shape a shared product vision?
We’re a four-person team that spans deep tech, research, product, and go-to-market. Two co-founders are AI engineers who’ve coded since childhood and hold AI degrees. One is our “human-in-the-loop” builder, obsessed with UX and interaction design; the other is a deep-tech architect who owns models, data pipelines, and scalability. Our CSO is a sustainability researcher with conference publications; she translates the latest science into defensible methods and metrics. Our fourth co-founder is a natural networker and storyteller who leads partnerships, pilots, and GTM.
Your platform uses multi-agent AI for personalised carbon insights. What sparked that idea, and how do you keep it both accurate and easy to use?
We went multi-agent because specialisation fixed our accuracy problem. One agent does analogies (turns CO₂ into simple comparisons), another does personalised recommendations, and separate document-ingestion agents power the business dashboard (they read bills, menus, invoices, utility data, etc.). By assigning each task to a specialist, we reduce errors and ensure more consistent and auditable outputs.
For users, nothing changes; they tap and get faster, more precise results. The extra complexity is hidden under the hood and requires a one-time setup on our side; after that, it runs automatically. In short, specialised agents mean higher accuracy for banks, clearer guidance for users, and no extra friction.
How do you balance building innovative tech with being mindful of its environmental footprint?
We balance innovation with footprint by following three rules: measure, minimise, and justify.
We track the energy/CO₂ intensity of our workloads per request and per feature release. If a feature can’t beat a defined carbon budget, it doesn’t ship.
We use specialist, lightweight agents; cache results; batch jobs; and run inference only when needed (no polling). We optimised for smaller/fine-tuned models, quantisation, and server autoscaling where possible.
Justify (net impact). Our agents replace hours of manual analysis with minutes of automated work and drive measurable emissions reductions for users and SMEs. We verify this with before/after baselines on activity data (transactions, invoices, utilities). If the abatement we generate doesn’t comfortably exceed our own compute footprint, we redesign.
In short, we make AI calls only when they add clear value, keep them as small as possible, run them as green as possible, and prove that the climate benefit to customers outweighs the cost.
How does AI help users understand and act on Scope 1–3 emissions? Any example where it led to real change?
AI’s real power here isn’t just automation, it’s personalisation. Think about it: Scope 1-3 emissions? For most people, that’s just jargon. A number on a screen means nothing. But AI digs into your spending patterns, your habits, your life. It learns what you care about. Maybe you’re a parent. Maybe you travel heaps. Maybe you’re obsessed with local food. Then it flips that carbon number into something visceral, your visceral. ‘Your monthly commute? That’s like melting X square meters of Arctic ice.’ Or ‘The emissions from your takeout habit? Equivalent to cutting down Y trees.’ This amount of carbon footprint could cause X amount of health issues in children. It’s not generic. It’s yours.
Then it goes further. Based on your behaviour: what you actually do, not what you say you’ll do, it serves up bite-sized actions it knows you might actually adopt. ‘Try this local grocer instead of delivery.’ ‘This bus route saves 30 mins and 2kg CO₂.’ And when you ignore it or engage? The AI learns. It gets sharper. It stops nagging and starts guiding.
You link carbon tracking with verified green investments. How do you make sure that’s both impactful and financially sustainable?
Honestly, we think carbon offsetting needs a fresh approach. The real change happens when we tackle both sides: demand and supply.
On the demand side, we're already working through bank APIs and smart behavioural nudges to help people understand and act on their carbon footprint. But we also want to revolutionise the supply side.
Here’s the exciting part: instead of just asking people to donate to offset their footprint, we want them to be able to invest. Imagine putting your money into carefully vetted green startups – businesses genuinely focused on reducing emissions. These startups get rigorously audited, so we know their real potential for impact and for solid returns.
This is a win-win:
For You (the user): You directly offset your carbon footprint while potentially growing your money. Its impact plus return.
For Green Startups: They get much-needed capital to grow, but only if they're truly building sustainable solutions. This is a powerful incentive.
For the Planet & Consumers: More capital flowing into sustainability sparks more innovation. This drives down costs and makes truly sustainable products and services more accessible and affordable for everyone over time.
So, by shifting investments towards verified green innovators, we're not just offsetting carbon – we're actively building a bigger, better supply of sustainable solutions for the future. It’s about making sustainability work for people and the planet, financially.
From student idea to full platform, what have been your biggest milestones, and what’s next for Carbon2Capital?
Turning Carbon2Capital from a student idea into a real platform has been quite the journey! Our biggest milestones? They’re all about resilience, relationships, and recognition:
Embracing the Grind (and the Criticism): Early on, we took our idea to incubators and experts... and honestly? We got a lot of tough feedback. It wasn’t always easy to hear, but it was exactly what we needed. Instead of getting discouraged, we dug in. We used every critique to sharpen our product and answer the hard questions. That phase didn’t feel glamorous, but it fundamentally shaped who we are.
Getting Our Foot in the Door (The UAE Programs): Getting accepted into the Emirati Tech Founders program (by SME & Bidayat) was a game-changer. Suddenly, we had dedicated mentors pushing us week after week to build, refine, and improve. It forced discipline and rapid progress.
Then came the NDTI program by Emirates NBD. This was critical. It plugged us directly into the banking world. We got invaluable access to people inside banks, learning exactly what they needed. This wasn't just theory anymore; it was about making sure our product met real banking requirements and delivered clear value for them. That first "foot in the door" was huge.
Going Global & Gaining Momentum: C2C is also chosen to represent the Emirates NBD at the London Tech Week in London, and now the EXPO event in Japan. This was an incredible honour and a major confidence boost.
Connecting with the Environmental Agency in Abu Dhabi opened important doors for future collaboration.
But the absolute pinnacle so far is The Hult Prize. Out of 15,000 startups globally, we made the Top 24. Now, we're in the running for the Top 7, living and working at their Ashridge residence, competing for that $1 million in funding. The energy here is unreal!
What's next for Carbon2Capital?
Winning the Hult Prize: Every ounce of energy is going into this final push for winning that game-changing $1M.
Launching Strategic Pilots: We're actively setting up our first live tests! This includes:
Piloting with a major UAE restaurant marketplace (stay tuned for the official announcement!).
Integrating with a key POS system provider to embed our tracking and nudges where spending happens.
Deepening bank conversations: Building on our NDTI connections, we're in active follow-ups with banks to secure our first major integration. (We know this takes time, but we're building the right relationships!).
These pilots are crucial steps. They get our technology into real user flows, prove value for partners, and generate the data we need to scale. It's about moving from milestones to impact – and we're hitting the ground running."
How did being part of DIFC FinTech Hive and the Hult Prize help shape your journey and scale your AI responsibly?
Honestly, DIFC FinTech Hive and the Hult Prize didn’t just speed us up – they fundamentally changed how we build. At DIFC, working with banks and regulators through NDTI was like getting thrown into the deep end of real-world fintech. They grilled us: ‘How’s your data protection?’ ‘Can your AI actually handle compliance?’ ‘What’s your fraud plan?’ That pressure was uncomfortable but necessary. It forced us to bake ethics into our tech from day one – strengthening anonymisation, adding transparency to our carbon algorithms, and designing real audits for the startups we onboard. No fluff. Just provable impact. They showed us how to integrate with banking systems safely because scaling responsibly means building something institutions can actually trust.
Then came Hult Prize. Competing against 15,000 startups? Brutal. Judges ripped into every assumption: ‘How do you prove this isn’t greenwashing?’, ‘What about global competitors?’, ‘How do you know people will reduce their footprint?’ To survive, we had to simplify everything. Trace every claim. Break down our logic. We audited our own carbon footprint, reworked our entire business model for clarity, not just for the pitch, but because scalable impact demands simplicity. And yeah, making Top 24 isn’t just about the prize money. It’s a stamp. Suddenly, partners actually listen when you say ‘Hult-backed.’
Together, these experiences rewired us. DIFC taught us to build responsibly inside the system. Hult forced us to think globally and be ruthlessly honest. Now, every time we tweak our AI, we ask: ‘Does this genuinely help people and the planet?’ That’s the filter. That’s how a student's idea grows into something banks and the Earth can believe in.
As you grow, how do you see Carbon2Capital evolving, especially around partnerships, impact, and inclusive sustainability?
Carbon2Capital has evolved a lot since we started, and honestly, we’ll keep evolving. We’ve even backtracked on some ideas when they didn’t pan out—but that’s how it goes in this space. New research rolls in, regulations shift, and we adapt. Our next big leap? Launching the investment side. Right now, we’re focused on the API, but soon we’ll connect users to vetted green startups. And as we grow into a more stable company, we’re committed to scaling impact globally, especially in developing countries. Climate change hits everywhere, even places without resources to fight it. That’s why we wrote a research paper outlining our audit methodology for these startups. One non-negotiable in that methodology is asking: 'Does this include marginalised communities?' Because real sustainability has to go hand-in-hand with accessibility. It’s not just carbon, it’s people.
What advice would you give other students or young professionals wanting to enter climate tech? How can they use their skills, whether in coding, audit, or design, to make a difference?
If you want to do something meaningful, entrepreneurship in climate tech is the way to go. It gives you a real opportunity to make a change. If doing work that has an impact makes you happy, do this. We all see climate change affecting us or people we care about, and it’s only getting worse unless we act. The tech we have now is cool, fun, and powerful – and as students, we have the tools to harness that power for good. Whatever you study – even arts – you can create a sustainable business model that promotes sustainability.
But be ready: it’s hard. Many experts or older professionals might not understand your passion. Climate tech has sometimes been used just for credibility, making people sceptical or even calling it a scam. You might struggle to convince them. So ask yourself: 'Is this idea something the world needs? Is it a viable business that can sustain itself, make money, and help the planet without harming it?' If you believe yes, stand behind it. We got tons of negative feedback early on. Push through. Those critics are smart people, but they don’t always see what’s possible.
Comments