AI/ML Opportunities in Nepal III

This article is the third 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. In part II of the series, I talked about AI/ML tech implementation in three business sectors of Nepal - the banking, insurance, and transportation sectors. Here in part III, I will go through the application of AI/ML in customer support, robotics, and vehicle traffic monitoring, all of which are of great potential for our nation.

Customer Support

A very large customer base means that no matter what you do, there will always be someone who is dissatisfied and has a bad experience. Although it may not have an effect on your bottom line right away, it chips away potential profits and mars company reputation. AI/ML-powered chatbots offer an inexpensive yet highly-effective solution. Even if you employ hundreds of customer service officers, they will continue to be swamped with queries and requests. Instead of customer support having to constantly deal with client issues all day, chatbots automate many of the simpler aspects involved in this so their time is more efficiently utilised. Using Natural Language Processing (NLP) plus other AI algorithms, Conversational AI is able to offer interactions between computers and humans that feel natural and human-like, without causing a jarring experience. This is for both chatbots and speech-enabled applications. Chatbots can automatically identify customer issues and intents, route or assign agents, allow for 24/7 service by providing customer service response suggestions, provide analytics using hard data for improvement, and save time and money by improving efficiency. In Nepal, chatbots are implemented in a minor capacity in some commercial banks such as Machhapuchchhre Bank’s “Maya” AI. These provide customers 24/7 access to information but are neither able to handle many things nor available in Nepali. Many improvements can be made. Businesses such as WorldLink can use existing data to create chatbots that identify problems using a preset options menu i.e. a menu tree. Service users can simply go through a few menus and submenus to receive specific instructions regarding their concerns and easily be directed to the right person. Chatbots will handle the repetitive questions and quickly route to the right person resulting in a much better customer experience.

Robotics

AI/ML implementation in robotics allows for a drastic boost in efficiency through automation. This can be through simple software routines without involving humans or packaging and other mechanical tasks including precision-assembly tasks. Nepal is understandably far down the ladder when it comes to robotics and its implementation. However, it is not out of reach. Naulo Restaurant in Durbarmarg is a fully digitized robotic restaurant that CNN covered in their news. Their sister company, Paaila Technology, developed these robots to serve tables as well as provide a good source of entertainment for customers, creating engagement and making the wait time feel lower than it actually is. Paaila Technology has also provided a chat robot named “Pari” to Nepal SBI Bank in Durbarmarg. Pari functions as an information desk and greets customers as well. It is the real-time equivalent of a chatbot and help desk.

Chatbots can automatically identify customer issues and intents, route or assign agents, allow for 24/7 service by providing customer service response suggestions, provide analytics using hard data for improvement, and save time and money by improving efficiency.

Through AI/ML implementation in robotics, small-scale agriculture can be improved through automated irrigation and sunlight or temperature control. However, a far more practical implementation is in weeding and crop care. Although interest in agriculture has been increasing, many Nepalis in villages leave for the cities and those in the cities head abroad in search of greener pastures. Manual labor costs continue to increase as well so there is great opportunity and profit in mechanization here. Large, commercial farmlands can employ one or two autonomous vegetable weeding robots such as Dino made by Naïo Technologies in France. This robot can cover almost up to 50,000 square meters of farmland in nine hours. Efficiency and productivity will go up whereas costs will go down. Utilising robots in the ways discussed above is a good place to start before we consider diving into the high-investment, high-profit venture of industrial automation.

Vehicle Traffic Monitoring

Last time, I talked about autonomous driving technology and how AI/ML can help in the trucking industry. However, that is mainly on highways, outside of cities. For urban areas, we can look at vehicle traffic monitoring through AI/ML implementations to greatly improve road traffic infrastructure. Currently, we have static traffic lights that work based on one-time surveys of traffic flow. Replacing them with dynamic ones will help alleviate many problems we are facing currently, especially regarding traffic congestion. With help from existing CCTV cameras and perhaps adding more so that there is one at every 500 meters or so, this enables proper identification and analysis of traffic. The data will let traffic lights be automatically managed by the AI system in real-time using data regarding average speeds, traffic flow, etc. The result is a smart traffic control system that adapts according to the current situation. It will be leagues better than the older model. An added benefit is the identification and management of rule-breakers. Since the CCTV cameras will be connected to the same data network, video evidence of traffic violations will be available. As they all connect to the computer vision tech, license plates can be detected and after human confirmation, the right vehicle can be registered if lane marking, traffic light rules, etc. are broken. Once vehicle owners are rightly identified using the mobile number registered with the number plate, traffic violators can then be contacted directly through automated SMS, informing them of their violation, fine, and how to deal with it. In the next part of the series, I will talk about AI/ML opportunities here in Nepal in matters of security, eCommerce, and personal financial management.