Carbon removal, built in the Global South. A company turning weeds, soil, and 170,000 farmers into credits that Google and Microsoft buy.
Above: the Varaha mark - seven leaves circling a gold sun. Named for the boar who, in the old story, lifts the Earth out of the water. The job description has not changed much.
Somewhere in Gujarat, a plant nobody wanted is being fed into a furnace on purpose.
In western India, a thorny invasive plant called Prosopis Juliflora is choking grasslands that used to feed cattle. Varaha harvests it, restores the land, and feeds the biomass into a pyrolysis facility. What comes out is biochar - a black, stable form of carbon that can stay locked in the ground for a thousand years or more.
That single project is why, in January 2025, Google signed what it called the largest biochar carbon removal deal to date: 100,000 tons, to be delivered by 2030. Not bad for a company that did not exist before 2022.
Today Varaha is a carbon project developer working across five countries - India, Nepal, Bangladesh, Bhutan and Ivory Coast - with about 170,000 smallholder farmers spread over roughly 1.7 million acres. It has removed more than two million tons of CO2 and issued around 150,000 verified credits.
It is also, unusually for a young climate startup, profitable after tax. The pitch is simple enough to fit on a seed packet: the next billion acres of carbon removal will not come from a lab in California. They will come from a farm in the Global South.
Carbon credits have spent years under suspicion. Too many projects promised removals that never happened, or counted trees that were never at risk. Buyers got burned. Headlines followed. The whole market wobbled.
Meanwhile, the people best positioned to actually pull carbon out of the air - smallholder farmers in Asia and Africa - were largely locked out. They lacked the satellites, the soil labs, the verification machinery, and the corporate contacts to sell a credit anyone would trust.
Varaha's founders looked at that gap and saw the whole business. If you could measure sequestration rigorously enough that a skeptical Google procurement team would sign off, and do it cheaply enough to work on a two-acre rice paddy, you could unlock an enormous, neglected supply of high-integrity removal.
The hard part was never the farming. It was the proof.
Varaha was founded in 2022 by Madhur Jain, Ankita Garg and Vishal Kuchanur. Their bet ran against the grain of climate-tech fashion, which at the time pointed at direct-air-capture machines in rich countries. They argued the opposite: that the cheapest, fastest, most durable removals were sitting in developing economies, waiting for someone to build the measurement layer.
Public face of the company and the voice on the Google biochar deal. Frames Varaha as proof that smallholders belong in the carbon-removal solution.
Runs operations across a footprint that now spans five countries and roughly 1.7 million acres of farmland.
Owns the technology - the remote sensing, machine learning and MRV platform that turns a field into a defensible number.
Varaha develops carbon removal projects across four pathways. Each is a different way to move carbon from the air into the ground - and each is only as valuable as the proof behind it.
Helping farmers stop burning crop residue and flooding rice fields, so soil holds more carbon and emits less methane.
Restoring trees on degraded and farmed land, turning marginal acres into long-term carbon stores.
Pyrolysis of biomass - including invasive species - into a stable carbon that can persist for 1,000 to 2,500 years.
Spreading crushed silicate rock on fields, where it reacts with CO2 and pulls it out of the atmosphere.
Underneath all four sits the part that actually matters: the MRV platform. Measurement, reporting and verification, built from remote sensing, machine learning and field science. It is the difference between a feel-good story and a credit a Fortune 500 buyer will put on its balance sheet. Credits are certified through registries including Puro.earth, Isometric, Verra, Gold Standard and Carbon Standards International.
Madhur Jain, Ankita Garg and Vishal Kuchanur start the company. A $4M seed round follows, backed by Orios, Omnivore, RTP Global and angel Kunal Shah.
Led by RTP Global, including Japan's Norinchukin Bank making its first-ever investment in an Indian startup.
Google agrees to buy 100,000 tons of biochar carbon removal - the largest such deal to date, delivered by 2030. Mirova later adds ~$30M to scale projects.
WestBridge Capital leads, in its first climate-tech bet. First $20M tranche funds expansion into Vietnam and Indonesia.
Skeptics welcome. That is roughly the whole point.
The customer list does the rest of the talking. Varaha's credits are bought by some of the most scrutinized procurement teams on earth.
Most carbon-removal money flows to expensive machines in wealthy countries. Varaha's mission is to flip that - to make high-integrity removal happen at planetary scale by paying the people who already work the land.
When a farmer in Nepal stops flooding a rice field or planting into burned residue, two things happen at once. The atmosphere gets a little cleaner, and the farmer gets paid. Climate action and rural income stop being a trade-off.
Return to that thorny invasive plant being fed into the furnace. A few years ago it was a nuisance - choking grassland, worth nothing. Now it is a contract with Google, a restored ecosystem, and a thousand-year carbon store. The plant did not change. What changed is that someone built the measurement, the market, and the math to make it count.
That is the quiet thing Varaha is really doing. Not inventing biochar, not inventing carbon credits - both are old. It is closing the trust gap that kept the Global South locked out of climate finance. As the company pushes into Vietnam and Indonesia, the same pattern travels with it: find the overlooked carbon, prove it rigorously, pay the farmer, sell the certainty.
The carbon market spent years being doubted. Varaha's wager is that the cure for doubt is not a better story. It is a better number. So far, the buyers who do this for a living seem to agree.