May 13, 2025.10 minutes

Dirt, data, and the dream: Cultivating hope (and a new economy) from the ground up

Nischal Harohalli Padmanabha

Let's be honest. "Climate change" can sound like a broken record, a scary movie stuck on repeat. Melting ice, rising seas – it feels huge, distant, maybe even... a bit abstract for daily life. But then you see the news: floods where there shouldn't be, fields bone-dry that used to be lush. It hits closer to home. This isn't just about polar bears anymore; it's about pasta on the shelves, water in the tap, and the air we breathe. It's about our own backyard, and how we messed it up by chasing progress without reading the planet's instruction manual.

And guess what one of the biggest culprits is? Our food. Yeah, the system that gets dinner to your table – from the farm to the fork – churns out a massive third of all those warming gases. We got here because for decades, the rule was "profit first, ask questions later." Industrialised farming fed a growing world, but it often forgot one crucial character in the story: the soil.

We hear a lot about cutting emissions, switching to solar, driving electric cars – all vital steps. But what about sucking the gunk out of the air? Turns out, we're standing on one of the planet's most powerful tools for this: healthy soil. Think of soil as a living, breathing sponge, eager to soak up carbon from the atmosphere and lock it away, building fertility and life as it goes.

The problem? Our traditional farming methods – the constant tilling, the heavy chemicals – have been like giving that sponge a hard squeeze, releasing the stored carbon back into the air and, worse, washing away the soil itself. Scientists warn we might have only 60 harvests left if we keep going like this. Sixty! That's not many dinners.

But there's a hopeful turn in the story: regenerative agriculture. It's farming that works with nature, not against it. It's about disturbing the soil less, planting cover crops, using natural fertilisers. It’s about letting the soil heal and do its job as a carbon-capturing champion. Imagine if we could get farms worldwide to switch? We could pull gigatonnes of carbon from the air!

Sounds great, right? So why isn't everyone doing it? Well, changing how you've farmed for generations is hard. It feels risky, might cost money upfront, and requires learning new tricks. Farmers operate on thin margins; asking them to take on this risk alone isn't fair.

Plus, there was a missing piece: proof. How do you prove that a farmer in Kansas or Germany is actually storing X tons of carbon in their soil? Without solid, verifiable data, banks won't offer green loans, big companies can't confidently invest in farm-based carbon offsets to meet their climate goals, and the market mechanisms needed to pay farmers for this vital work simply don't materialise. We were stuck. We needed data, and we needed it at scale.

This is where our journey really started. We looked at this massive challenge – the climate crisis, the stressed farmers, the incredible potential of soil – and we had a vision: What if we could use technology to bridge that gap? What if we could make the invisible work of carbon sequestration in soil visible, measurable, and valuable?

We knew sending people with shovels to every field wasn't the answer. But overhead? That was a different story. Satellites constantly circling the Earth collect mountains of data – pictures and spectral information showing what's happening on every acre. Could we teach computers, using AI and machine learning, to read those fields like a book? To identify cover crops and crops, detect tillage, and ultimately, estimate how much carbon was being stored?

Turning that vision into reality wasn't like flipping a switch. It was like building a complex machine piece by piece. Our initial strategy (v1) was about getting our own house in order. We had talented people, but we needed to work together seamlessly, like a well-oiled farm team.

Our strategy was guided by simple, yet powerful, principles:

• Data is gold: Treat the information from satellites and farms as our most precious asset.

• Share the gold: Make that data accessible and understandable to everyone who needs it, inside and outside the company.

• Talk and try: Encourage collaboration between scientists, engineers, and business experts, and don't be afraid to experiment.

• Build trustworthy tools: Our AI had to be reliable and fair – no black boxes.

• Learn constantly: Keep checking our work and improving.

• Think "platform": Build shared tools and processes so everyone isn't reinventing the wheel.

Our execution involved reorganising our teams, getting clear on our goals, and starting to build the foundational data systems. It was about creating the internal strength needed to tackle the immense external challenge. We set clear goals – faster model deployment, better data access, becoming leaders in using satellite AI for farming. We were building the bedrock.

But here's another crucial part of the story: it's not enough just to build a system that can process data fast. If the "smart computer brains" (the AI models) doing the analysis aren't reliable, aren't accurate, then scaling is not valuable. It's like having a super-fast printer that just prints random letters.

For us, with the goal of creating high-quality carbon credits, the reliability of our models is paramount. Trust is everything in carbon markets. So, we put our machine learning models through intense testing and development.

Think of it like detective work combined with meticulous craftsmanship. Our foundation wasn't just code; it was built on terabytes of real-world evidence: ground truth data collected over several years, across different seasons and countries, carefully annotated by human experts. This gave our AI a solid base of understanding.

Using this foundation, our team invested thousands of hours, like skilled artisans, in building our deep learning models. We focused on creating models we could truly understand and, if needed, modify. This wasn't about building a mysterious black box; it was about knowing why the AI was making its predictions. This deep understanding and the rigorous evaluation process give us confidence in the quality of our predictions when we finally run them at scale across millions of hectares. Reliable models are the beating heart of our credibility.

Then came the next chapter of our execution: building the engine that could process data not just for a few farms, but for entire regions - even continents! Our old system was like trying to water a million fields with a teacup. We needed a monsoon.

Our strategy evolved. We realised processing individual fields was too slow. The satellites see the world in big squares (tiles). So, we shifted to processing those whole squares at once. It was technically challenging – huge amounts of data! – but it was the key to scale.

We brought in powerful tools like Ray and Anyscale, like adding turbo boosters to our engine. This wasn't just a tech upgrade; it was about empowering our scientists. Instead of getting bogged down in complex coding for massive computers, they could focus on the science, on building better AI models. The platform handled the heavy lifting.

The results were incredible, a testament to focused execution:

• Speed of light: Processing data for millions of hectares went from taking months to taking one hour. That's a 10,000x leap!

• Smart savings: We slashed infrastructure costs by 78% – being efficient with computing power is also good for the planet's energy bill.

• Faster innovation: Getting a new AI model from the drawing board to real-world use went from months to weeks.

• Instant answers: Our users and partners can now get detailed insights on fields in seconds, not hours.

This platform isn't just code; it's the engine that generates the proof – the credible, verifiable data needed to show exactly how much carbon is being sequestered, field by field, year after year.

We built the engine, and it's running strong! Our initial strategy (v1) worked. But the journey continues. Our vision is bigger – more fields, more insights, more impact. Our current strategy (v2) is about refining, optimising, and reaching even higher. We're focused on being faster, higher, stronger. Building on the strong foundation we've laid, we are now in a position to dramatically foster innovation with AI and bring higher-value models into action, taking on problems that very few dare to tackle.

This means making our internal processes even smoother, developing better ways to measure our impact in the real world (how are those reliable models performing "in the wild"?), and making sure everyone in the company can easily access and use the wealth of data we're generating. It's about continuous improvement, guided by principles like:

• Team playbook: Standardising how we build things so everyone can contribute easily.

• Knowing our score: Tracking key metrics to see how we're doing and where we can improve.

• Eyes everywhere: Building systems to spot potential issues before they become problems.

• Growing our people: Investing in our team's skills and knowledge.

And supporting this are our core platforms – the refined ML engine, a smart system for managing all our diverse data (like a well-organised library), and a new platform specifically for turning scientific discoveries about carbon into production-ready tools. These are the highways for our future innovations.

With this verifiable data, we're creating the foundation for that new economy. Banks can now see the proof and offer financial products that support farmers transitioning to regenerative practices. And, crucially, that data turns carbon sequestration into a valuable asset – a carbon credit.

These credits are like certificates proving a farmer has stored carbon. Companies looking to reduce their environmental footprint can buy these credits, providing farmers with a new income stream. It's a direct financial reward for being a steward of the land and fighting climate change.

The power flows back to the farmer: extra income from carbon credits, access to green financing, and digital tools giving them unprecedented insights into their own land. They become active participants and beneficiaries in the fight against climate change, not just observers.

Yes, the climate challenge is huge. The old narrative can feel overwhelming. But look closely, and you'll see countless stories of innovation and hope being written, driven by technology, clear vision, smart strategy, and dedicated execution.

We've gone from a necessary idea to a scalable platform, enabling a new economic model that rewards healing the planet. Billions are flowing into this space, attracting incredible talent combining tech, science, and business know-how.

We've built the engine, charted the course, and are actively cultivating this greener tomorrow, one field, one data point, one verifiable credit at a time.

Because ultimately, this is about more than just carbon numbers. It's about ensuring a future where agriculture thrives in harmony with the planet, providing food and a stable climate. And as the saying goes, it's the only planet we know of with beer. So let's keep building, keep growing, and keep hoping. The dirt, the data, and the dream are showing us the way.

Learn more

Explore our latest articles

Unearthing the potential: What is soil, and why does it matter in the voluntary carbon market?

Read the full story

From credit to credibility: Building a high-integrity soil carbon project

Read the full story

A message from our CEO: Agreena's Verra registration is a milestone for regenerative agriculture

Read the full story