It’s Tuesday, and today we’re diving into BemAgro, a Brazilian agritech startup. Founded in 2018 by Johann Coelho, the company just completed a $5.8 million Series A round led by The Yield Lab LATAM, with Colorado Ventures, CNH, Atvos, and Agroven also participating.
The Context
Today I’ll talk about Brazil’s agriculture. To understand how BemAgro fits into the industry, I want to focus on two core points: how massive the sector is and how technology has shaped it.
Agriculture’s Size And Role
Brazil is known as an agricultural powerhouse. The sector accounts for 16.2% of employment and 40% of total exports. It contributes 8.4% of GDP, and the agrifood sector as a whole adds 22% to the country’s GDP. Brazil is the world’s leading producer of sugarcane, soybeans, and coffee, among other crops. Aside from its historical role and climate variability, the biggest reason Brazil plays such a prominent role in global agriculture is simple: it is a huge country, with an area of 8.5 million km2. Soybean planted area alone is estimated at 470,000 km2. All that land has to be monitored and managed. There are several layers to this.
Area size. The sheer scale creates management costs. Someone has to monitor these vast sowing areas which, depending on the region, can be complicated by challenging terrain. Many crop-growing areas are controlled by large holdings. Some agro-industrial groups manage tens of thousands of square kilometres of fields. It is hard and expensive to monitor 100% of these areas, so companies often rely on sample-based monitoring, which reduces accuracy.
Cropping intensity. A practice known as safrinha is common in Brazil. Corn is planted immediately after harvesting a primary crop, usually soybeans. This practice accounts for 78% of total corn production. It increases monitoring and execution pressure because there are multiple critical planting and harvest windows per year, not just one.
Source: Farmdoc
Chemical components. Since arable areas are massive, crop chemical market is also massive. In 2022, the market was estimated to reach $21.7 billion. On top of that, with 85% of fertilizers imported, the sector remains highly exposed to exchange-rate risk.. A weaker real increases input costs. Either these costs are swallowed by producers, which compresses margins, or they are passed through, which makes Brazilian agricultural exports become less competitive. Thus any consistent savings in chemical-related expenses would directly contribute to the bottom line and/or sustain Brazil’s competitiveness on the global stage.
Technology Contribution
I usually write about the agrisector in countries that lack basic technology-adjacent tools. That was the case with Tanzania, Kenya, and Indonesia. That is not the case with Brazil.
First, the sector is mechanized. In 2023, there were 1.35 million tractors in operation. Since 2015, most tractors sold in the country have been equipped with GPS and telemetry.
Second, mapping data is digitized. Brazil has strong public and semi-public geospatial data infrastructure. One example is MapBiomas, which offers multi-institutional land-use and land-cover maps.
Third, farmers have adopted precision agriculture. 84% of farmers report using at least one modern technology, such as GPS, in their production process. Among those using smartphones for agricultural purposes, 71% rely on specific apps to access crop information or monitor pests.
Precision and digital agriculture technologies used by farmers. Source: Bolfe et. al.
Yet technology remains underutilized. Over 80% of farmers do not access or use data gathered from on- and off-field sensors, and most digital systems embedded in tractors are not actively used.
Which brings me to BemAgro.
The Product
BemAgro provides an AI-driven platform for agricultural businesses to plan, monitor, analyze, and adjust how they manage their fields. The core value it offers is improving efficiency at every step of the agricultural cycle, from planting to harvesting. The company does this by building and maintaining a regularly updated corpus of data from satellites, drones, and agricultural machines.
That data is then used to model 180 planting scenarios, all aimed at helping farming businesses reduce input consumption and optimize machinery and labor use. By implementing BemAgro, producers can lessen pesticide use by up to 80%, cut the number of machinery maneuvers by 25%, and improve planting density by up to 10%.
Now, let’s get into how all that happens.
While the company defines four core modules, I would group its activity into three main stages:
Preparation. This is the initial analytical layer built on the data sources mentioned above. The platform extracts machinery data to construct altimetry and topography maps, processes satellite and drone images to generate detailed field maps, and collects vegetation indices to assess crop variability. At this stage, the goal is to create a granular digital twin of the field before any major operational decision is made.
Planting. Based on soil type and preparation, rainfall index, slope and drainage, and equipment dimensions, the platform generates 180 possible scenarios. These scenarios inform terrace line design, planting lines, and spraying lines. In practical terms, this defines how the entire field will be structured and how planting and spraying operations will be executed, down to spacing and route optimization.
Monitoring. After planting, the platform shifts to management. Farmers can analyze where weeds may appear and what types; identify areas with variable growth and then instruct machines or spraying drones to apply growth regulators, urea, or other chemical inputs only where needed; detect planting gaps through plant counting which, in the case of coffee, could mean addressing a 10% gap that translates into 1.5 years of lost harvest over a 15-year cycle; and analyze parallelism to ensure GPS-guided machines do not overlap passes or run over crops.
A similar solution set has been developed for forest management. There, the platform supports soil preparation and planting activities, measures seedling survival and planting gaps, and enables residue mapping and related calculations.
The platform is accessible through both a web version and a mobile app. The latter allows for offline map usage, evidence recording, and field navigation, which is critical in large rural areas with limited connectivity.
Source: Google Play
The Business Model
Three elements are at the core of the model:
Low CAPEX. BemAgro requires no hardware investment from the client. Because the platform is compatible with over 40 tractor brands and works with data from drones and satellites already in use, farmers can extract value without installing new telemetry systems or sensors.
Asset-light strategy. The company is 100% asset-light, meaning they do not own or operate any drones or machinery. This allows them to focus exclusively on Al processing and computer vision, delegating data collection to the clients or their service.
Investors = clients. The company targets large agro-industrial groups, while also reaching smaller farms (often lacking specialized labor) through a distributor network. But what’s fascinating is that the large groups aren’t only buyers, but financiers. Atvos (ethanol) and Suzano (paper) have invested and signed multi-million-dollar contracts, which effectively does two things at once: it injects cash into the business, and it turns customers into stakeholders who have a direct incentive to push adoption internally and stick with the platform over time.
Taken together, these elements create resilience. With no CAPEX required, the barriers to adoption are relatively low. An asset-light structure allows the company to expand or reduce coverage and introduce or adjust modules without being constrained by physical infrastructure. Having strategic clients on the cap table makes it easier to invest ahead of demand, since key customers are directly incentivized to support platform improvement and integration.
Finally, BemAgro is working to ensure long-term client generation. There is a shortage of professionals capable of reading and interpreting complex agricultural data. To address this, the company launched BemAgro Academy, a free online platform that has already certified hundreds of professionals in geotechnology and AI. These individuals often go on to work for agricultural operators, accelerating digital adoption in the sector. Of course with the help of BemAgro.
Monetization
The company monetizes through software subscriptions.
Results
The platform has processed 7 million hectares, serving over 400 companies in 11 countries. BemAgro has~tripled~ its revenue over the past 1.5 years.
The Bear Case
The biggest threat to me is on-the-ground execution. Computer-vision outputs may very well be directionally correct, but still encounter issues: operators can make mistakes, imagery timing may be incorrect, etc. If results are inconsistent in practice, customers may start viewing the product as useful, but not critical to their success.
Two additional concerns give me pause. First, there is concentration risk. If a meaningful share of revenue comes from a handful of large agro-industrial groups, any disruption in their industry, whether commodity price shocks, regulatory changes, or input volatility, can impact BemAgro’s revenue. And I also wonder, how expensive is the compute, whether it benefits from scale economies as volumes rise, or whether costs will simply track data inputs one-for-one as more data flows in.
The Bull Case
If BemAgro delivers results reliably, it becomes an integral execution layer and a part of the infrastructure. Farm operators embed their product into their season planning, increasing the switching costs as the platform gathers more data and trains on specific regions. Equipment providers see how the ROI of their machinery rises, creating more loyal customers, which pushes them to develop the distribution partnership. This is a rare time where both the client (farmer) and the supplier (equipment manufacturers) can take advantage of higher a ROI by implementing an intermediary (BemAgro).
The second part of the bull case is the customer-investor dynamic. It provides revenue and validation at the same time, making customer acquisition easier for BemAgro and harder for smaller competitors to replicate. As these clients become repeatable references and distribution channels, BemAgro becomes unusually efficient at acquiring new customers.
The Takeaway
What’s the one lesson investors and founders can take away from BemAgro?
Converting clients into investors is an underutilized capital-raising tool. In emerging markets, where capital is scarcer, allowing clients to invest solves two problems: access to capital and proof that the product works. Obviously, this only works in industries where your main clients are large enterprises. But still.
