Agriculture is one of the world’s oldest professions, tends to rely on traditional, outdated workflows, and has historically resisted innovation. But with many countries nearing the brink of a major food crisis, failure to optimize agriculture could have serious consequences.
For example, Brazil has one of the world’s fastest-growing agricultural economies, but its farming practices tend to be less efficient than other countries. There are many reasons for this shortcoming, but two of them are the slow integration of new technologies and the difficulty of receiving financial assistance to fund such innovations.
Abhay Sarda of agricultural solutions company Traive is addressing both these issues by using machine learning to automate farmers’ financial management and crop monitoring.
Brazil’s Productivity Gaps and Growth Potential
Brazil is one of the world’s largest agricultural producers, exporting around 522 million tons annually of some of the world’s most important agricultural commodities.
But even though Brazil has the most arable land in the world, it lags behind the United States in agricultural productivity, measured by yield, production, and the efficiency of resource use in production. In fact, Brazil’s productivity is about half that of the United States, according to Abey.
So why can’t Brazil’s agriculture sector develop further? The country certainly has the potential: its favorable soil conditions and low resource costs allow it to grow more crops on more land (and at lower cost) than any other country in the world.
One obstacle is the prevalence of outdated agricultural techniques: traditional methods of harvesting crops are not efficient enough to meet growing market demands or adapt to a changing climate. This is partly due to the second obstacle: limited access to financing for Brazilian farmers. Diminishing low-interest credit for essential loans prevents farmers from investing in modernizing their operations and optimizing production.
Abey Sarda and Treive appear.
How Abey Sarda is closing the gap with machine learning
Abhay Sarda became interested in data science while writing his thesis on improving solar energy efficiency at the National University of Singapore, where he saw first-hand how machine learning could approach complex problems like energy forecasting. This led him to pursue a master’s in computer engineering at Boston University, where he met his co-founders of Traive.
At the time, Traive was a three-month-old startup with two employees and no funding. Joining this growing company was a risk, but Abhay wanted to build something from scratch rather than simply follow the path of his peers who were taking jobs at large tech companies. Five and a half years later, Traive has more than 100 employees across Brazil and the U.S., and Abhay is director of machine learning, overseeing the company’s data and AI strategy.
AI-based credit risk assessment
At Traive, Abhay uses machine learning to help clients better understand the risks of lending to farmers. Using a ground-level sensor network and cutting-edge satellite imagery, Traive’s algorithms can perform risk assessments at scale, from crop monitoring to yield forecasting.
Using these technologies, Abay has pioneered the development of an automated credit scoring system that has processed 10,000 loan applications from farmers over the past few years, amounting to 10 billion Brazilian reals, or approximately US$1.82 billion, and the company plans to quadruple this number over the next two to three years.
Today, the company has one of the largest databases of customer records in Brazil’s agricultural sector, and many of the country’s leading enterprise resource planning (ERP) providers have integrated Traive’s risk assessment tools into their solutions. As a result, Traive has become one of the leading credit assessment tools in Brazil, providing an innovative and reliable solution to the 47% of Brazilian farmers who need to apply for high-risk agricultural loans from commercial banks.
Intelligent Crop Monitoring System
One of the foundational elements of Traive’s industry-leading risk assessment tool is advanced crop monitoring technology. After all, farm size, crop condition, and irrigation needs are all factors that help investors understand which projects are more likely to succeed. Using Bayesian methods to analyze detailed data specific to Brazilian agriculture, Traive uses AI to parse historical and contextual data to create advanced models that can predict future outcomes.
This not only brings clarity and transparency to investors, so they know what their investment money is going towards, but also provides a range of benefits to farmers. For example, an intelligent crop monitoring system can detect problems such as pest infestations or nutrient deficiencies early, allowing farmers to act in time to protect their crops. It can also provide accurate information on nutrient levels, soil moisture and up-to-date weather conditions, reducing overuse of resources such as fertilizer, which is a scarce resource in Brazil.
Abhay’s efforts have enabled Traive to better support agricultural projects that are not only profitable but also environmentally friendly. The results are clear: To date, Traive has helped finance over 10 million hectares of soybeans, 4 million hectares of corn, and 1 million hectares of cotton, all while leveraging the power of data to help enforce environmental laws.
A new era of agricultural business solutions
Abey Sarda believes that by using machine learning to simplify loan applications for farmers and improve agricultural practices, Brazil can close the productivity gap with the United States and realize its full potential as an agricultural powerhouse.
Not only does Traive represent a new era for agriculture in Brazil, but Abhay believes that with AI as the industry’s co-pilot, farmers can meet growing market demand for their produce and ultimately make great strides toward preventing a global food crisis.
Spencer Hulse is the editorial director of Grit Daily, where he oversees other editors and writers, as well as day-to-day operations and breaking news coverage.
