AI & ML Opportunities in Nepal

This article is the first in a series where I explore the implementation of Artificial Intelligence and Machine Learning technologies in Nepal. Some of these can include businesses that are working in this space, based out of Nepal. If you are unfamiliar with the term AI/ML, it stands for Artificial Intelligence/Machine Learning. Machine Learning involves large datasets that use neural networks to learn and the output is usually for an Artificial Intelligence system.

How is AI/ML relevant for Nepal?

Most traditional Nepali businesses still do not have a proper functioning website. Digital marketing for many is still making posts on Facebook and running competitions to increase their page likes. Keeping this in mind, are businesses going to understand and be able to leverage AI/ML in Nepal? A simple implementation of this technology is something that most of us who use email are probably familiar with already: the spam filter in your email incorporates AI/ML. Using tons of data from spam mail collected throughout the years, email providers such as Google have a built-in spam filter. Even such a small thing that we take for granted, which is actually quite helpful, exists because of the advancement of this tech. I want to examine three traditional industries in Nepal, where AI/ML can be used.

The Agriculture Sector

Microsoft and the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) have been working in Andhra Pradesh, India, and helping local farmers reduce costs and increase production. Together, they developed a sowing date application that uses cloud-based predictive analytics to advise farmers on the ideal time to sow their seeds. This, of course, took into account various factors such as weather conditions and soil type. Microsoft India claims that this has resulted in a 30% higher average yield per hectare. We can adopt some form of “precision agriculture” here in Nepal similar to what is being done in Andhra Pradesh. For example, we can start small with simpler economic sensors that measure soil humidity, salinity, and nutrient content. This way, farmers can actively gauge growth conditions and crop development much more effectively. As the profit margin increases, more investment can be made so that the output increases with the same or perhaps lower input.

The Education Sector

Due to the current lockdown, many schools and colleges that are able to teach their students online have switched to an online instruction delivery model. However, replicating in-class instructional format delivered via video calls does not truly utilise the strengths of an online education system. It is possible to have an education software where you have all the standard functionalities in addition to an AI system which automatically identifies the core concepts that an individual student or the entire class is struggling with and can provide additional instructional and practice materials to teach that concept without any human intervention. Such a system has been built by Fusemachines, a NY, USA based company founded by a Nepali with significant operations in Kathmandu. With tech such as this, education can be provided and student progress can be monitored even in the midst of a nationwide crisis. The use of AI-enabled systems in education is just starting and we can expect to see more of similar systems implemented in Nepal as soon as educational institutions are forced to adopt new ways of teaching students.

The Health Sector

Experienced/skilled doctors are naturally better at providing medical diagnoses. Unfortunately, distant parts of the country lack such professionals as we are well aware of the acute shortage of doctors in remote government hospitals or health posts. AI has been getting better at detecting diseases through medical imaging, with their accuracy coming at par with healthcare professionals. This can help with disease detection for people who can get medical imaging done, or can be used as a form of getting a second opinion by confirming if the diagnosis of a healthcare professional is accurate. CloudFactory, another company with a significant presence in Nepal, collaborated with V7 Labs in London to contribute to Covid 19 research. They released annotated X-ray datasets allowing for radiologists to distinguish more quickly whether the patient has Covid 19 pneumonia specifically or some other ailment. By applying machine learning to identify the X-rays of patients, the analysis time is cut short, leading to a faster diagnosis. All three sectors that I explored above are much more vast and have multiple applications of AI/ML technologies. The ones that I mentioned are ones that are already being done either in Nepal or elsewhere, and are relevant for Nepal. Additionally, we already have human resources here who are capable of implementing such solutions. AI does not always have to depend on ML models but the ones that do need a large, diverse, high-quality data set and that is where a lot of the latest implementations are being done. In the next part of the series, I want to have a look at the banking, insurance, and transport industries in Nepal regarding the way they can leverage AI/ML to further their development by leaps and bounds.