Interview with Diego Saez Gil, Cofounder and CEO of Pachama

Forests play a vital role in combating global warming. Today, we meet Diego Saez Gil, Cofounder and CEO of Pachama, a San Francisco-based startup leveraging data, artificial intelligence, and automation to protect ecosystems, restore forests, and improve carbon markets.

To go back to the very beginning, I always like to share that I grew up in the North of Argentina in an area with beautiful forests and nature around me. I think that this made me, as a kid already, some kind of ecologist and someone deeply in love with  nature.

But then I kind of forgot about this life, got a job, and focused on making money and making my place in the world. I moved to Europe, to do my masters in Barcelona, then went to New York to do my internship. Then, I started my first start-up, a mobile app to book hostels, I had no idea of what I was doing. It was very difficult at the beginning but somehow a few years later we managed to get an acquisition offer by a big travel company.

After that, I started Bluesmart which was a smart luggage company. The idea came when I lost my suitcase and I was wondering how come there is not an easy way to track your luggage. We came up with a prototype, we launched a crowdfunding campaign and entered the race of building a hardware company. As you know, building a hardware company is extremely difficult, you have to go to China to set up the manufacturing, you have to set up the logistics and distribution around the world, you have to be constantly innovating in both hardware and software. Fortunately, we had a very innovative product and things were going great, except that, at some point, airlines decided to ban built-in lithium batteries because some samsung phones had caught fire on airplanes. That was totally unexpected and led us to sell the company very quickly, mostly thanks to the patents and the IP we had, but otherwise it was really a crashing situation.

After that I decided to take some time to reflect. I had been running and running for many years. During that time of reflection, I started reading about climate change and got really concerned about the velocity in which we were running towards a catastrophe if nothing was done. Secondly, I reconnected with the love for nature that I had from my childhood. I realized that I could put everything I had learned about building tech companies at the service of restoring nature to solve climate change. That is the mission of Pachama: restoring nature to solve climate change.

During the building of these 2 companies, especially Bluesmart, I became very familiar with the state of the art of remote sensing, satellites data, AI, big data and many other technologies that can be used to better measure forest, estimate carbon sequestration, and therefore to bring more transparency to the carbon market. That is the core of the idea behind Pachama.

We need every company and government to achieve net zero carbon emission. That means that these organisations need to measure and reduce their carbon footprint, moving away from fossil fuels. But, while they do that, which may take 1 to 2 decades, they need to start compensating their emissions that they keep putting in the atmosphere.

That created a market for carbon removal or carbon sequestration that is driven by carbon credits, which is a unit that represents that someone somewhere has removed a ton of carbon from the atmosphere. One of the types of projects that can receive carbon credits is reforestation or forest conservation. Until recently it was difficult for these companies to send money to another country where someone was protecting or restoring forests, because it was a leap of faith. While there are certifications to issue these credits, it is not very easy to track these projects. So, what we came up with is the capacity to produce data in real time, for monitoring and validating how much carbon these forests are sequestering in real time and if these projects are delivering on their promises.

Then, we realized that another problem with this market was that you have projects in different countries like Brazil, Peru, Indonesia, Congo, Nicaragua and you have all these offset buyers in Europe, US, Canada, China and it’s difficult for the parties to connect. This was a typical case for an online marketplace. So the end result is an online platform that connects supply and demand, and in the middle, uses data to verify and monitor the impact of the projects that are being funded by these companies.

I like to say to investors that this market is either trillions of dollars or the planet is screwed. There are no other scenarios. Either in the next 2 decades we drive trillions of dollars from the polluters to the solutions to climate change, or our civilization is up for big trouble. And it is a possibility that this market does not take off and we will have to face the consequences of climate change.

I am hopeful that the market will continue to grow at the current speed and that by 2040 we’re going to have hundreds of billions dollars flowing to carbon removal projects.

I’ve taken those numbers from a World Bank’s report and that includes compliance and voluntary markets. Some of these compliance markets are government to government agreements. And yes, it used to be the case historically that this fundings would go for energy efficiency or for capping off emissions by a polluter.

But now, we believe that that money needs to go to carbon removal and we particularly believe that it should go to nature carbon removal because nature carbon removal is so efficient. When you’re restoring a forest, you are not only removing CO2 from the atmosphere, you are also creating habitats for endangered species, you’re conserving biodiversity and water, and you are creating local jobs for communities.

That’s right. Today, there are other explorations of how we can remove carbon from the atmosphere that include building machines that suck up air and convert it into a material or filters that you can put in factories to capture carbon before it goes out, or weathering of minerals on the beach to produce a chemical reaction and then sequestrate carbon.

All these are interesting areas that need further exploration but we have these other technologies that have been under developpement for billions of years that work very well. You can get out 200 hundred tons of carbon in a single hectare if you restore a tropical forest. The Crowter lab from ETH Zurich University led a study in 2019 and they showed that there was 1 billion hectares available to plant trees, which can support up to 1 trillion trees. So if you do the maths, that is 200 gigatons of carbon that could be captured thanks to this billion hectare, that is ⅔ of the total emission that we have produced since the Industrial Revolution.

You may wonder why we haven’t already driven more money in that solution if it has such a high impact? Well, there are challenges, measurement & monitoring that we’re trying to tackle at Pachama, then scaling reforestation efforts and generally the bottleneck is producing seedlings and saplings in nurseries at a speed that is enough to reforest thousands of hectares. But frankly, that is a problem that can be solved and there are several organizations that are working on this. We humans are capable of doing such incredible things, we’re building big bridges, tunnels, massive buildings, we mine giant mountains. How come we cannot plant a trillion trees? When we put our engineering efforts into a challenge, we solve it. So, there are challenges in scaling reforestation but I think we can solve them.

We use deep learning, convolutional neural networks to analyse and create models that predict things coming from satellite images. So, for example, we’ve trained a model that combines satellite data with lidar data coming from airplanes, and ground data collected by foresters. Our deep learning algorithms get trained to be able to make predictions on how much carbon is there on a given batch of forest.

We built another model in with which we predict future deforestation and we can use that as a baseline to predict what’s going to happen if nothing is done in the forest. Let’s take the Amazon rainforest for instance. Deforestation happens in a quite predictable pattern there: a new road is opened and, around it, smaller roads get opened, and then deforestation starts to happen and expands all around these roads. And when you look at historical data that we have from 1984 until today, you can see how deforestation has been expanding in a region.

And then another machine learning model that we are building is a model to monitor and detect disturbances in the forest in real time. We pull radar data and compare images from 2 different periods, detecting changes and we have an algorithm that alerts us, we clean the false positives and then we investigate the true positives.

Yes, we are making good progress on transparency, we have a lot of our projects on our website, where you can go and see the maps. We are adding more and more layers of data on top of the maps, where we show where the carbon is concentrated, where the deforestation is happening, the charts of biomass change over time, etc.

Hopefully in the future we can also provide data for country level and we can look for example at Brazil and Peru, and see what is going on in the forest, where the forest is being deforested and where the forest is growing.

One of the main focuses this year is to use this AI and this data to help originate new projects. Today, there are a lot of existing projects that are selling carbon credits but if we are to solve climate change we need to start a lot of new reforestation projects around the world. We are going to use this data to prequalify areas where projects could exist, and then partner with landowners and organizations on the ground that could get started with reforestation and we’re going to help them accelerate towards certifying carbon credit and going to the market.

We’ve announced our recent project in Brazil in the Atlantic forest in partnership with Mercado Libre which is one of the largest tech companies in Latin America.

And, yes in terms of hiring we continue growing our team, from 19 to nearly 30 in 2021. We’ve just hired people from Microsoft, Facebook, Open AI and Tesla. We want to hire super passionate and smart people who want to work in climate change, machine learning engineers, forest remote sensing specialists, and product managers as well.

I like to think that the purpose of AI is to protect planet Earth. I think that in the future, if everything goes right, we will look back and realize that we have developed all these computing powers so that we could analyze and protect our own planet. So what gets me excited is all the data we’re going to be able to produce, not only from satellites but data coming from sensors, all sources, DNA data about the forest. Imagine you can know everything about all the ecosystems that make up our planet and then use data to protect it, to ensure the long term existence of civilization and the other species on our planet.

It is not going to be only Pachama, it’s going to be many organizations, some focusing on biodiversity or ocean. For instance, Planet has satellites and is investing a lot in ecosystem analysis, there is another startup in Europe called Overstory that is also analysing forets for fire prevention. The Xprize has also launched a competition around rainforest and biodiversity analysis, there are a few teams competing there that are super interesting.

It takes a village really. Sometimes it drives me crazy to see so many brilliant people working on making people clicking on ads, working for the military, for mining and oil. I hope we’ll manage eventually to gather all these brilliant brains to focus on protecting our planet.