Kamal KC
In recent times, artificial intelligence (AI) has been gaining global attention. AI is increasingly being integrated into every service and product. However, ten years ago, few would have predicted that AI would gain such widespread popularity!
Even Dr. Suresh Manandhar, who completed a PhD in AI from the UK three decades ago, hadn’t expected AI to become popular so quickly. “I didn’t think AI would arrive this fast. I thought it would take another 5–10 years. But it only started gaining attention after the arrival of ChatGPT,” Manandhar says. “Fifteen years ago, even those who studied AI found it hard to get jobs.”
Forty years ago, when Suresh said he wanted to study AI, many people questioned his decision. But today, many of those same people are using AI in one form or another.
A Passion for Experimentation
Born in Kathmandu, Suresh Manandhar wasn’t particularly focused on school studies. However, he had a deep interest in experimentation and self-learning. His room was often filled more with library books and experimental gadgets than textbooks.
Suresh recalls that his room as a child resembled a laboratory. “There was an American Library in Khichapokhari. I developed an interest in electronics from reading there,” he says. “I used to build and break model airplanes, motors, and wire-based devices. I loved trial and error. My room was like a lab.”
Back then, he was especially passionate about airplanes and built many model planes at a time when model plane kits were hard to find—so he had to build everything himself.
His parents, seeing his passion for technology but not for academics, doubted if he’d even pass the SLC (School Leaving Certificate). However, he passed with second division. He then went to India for higher secondary and engineering studies.
He completed his bachelor’s degree in Electrical and Electronics Engineering in India. During his final year, he became interested in three subjects: power transmission, digital control systems, and programming. He completed his thesis on programming combined with hardware.
Back then, computers weren’t as accessible, and he had to use assembly coding. After graduating in late 1992, he returned to Nepal. The first thing he asked his father was to help him find a place in Nepal where he could work with computers.
At the time, the National Computer Center at Singha Durbar had computers. His father, who was the librarian administrator at Tri-Chandra College, spoke to the center's director.
After his father arranged a meeting, the director invited Suresh for an interview on a Saturday. He began working the next day as a system engineer.
As he worked, he realized he couldn’t remain confined to just that center. So he started visiting the American and British libraries again. It was during these visits that he learned about AI, and his interest in the field grew. By then, he had already decided not to pursue further studies in electrical and electronics.
Once his interest in AI deepened, he went to the UK on a scholarship for a master’s in AI. In 1987, he completed his master’s from the University of Essex and returned to Nepal.
Back in Nepal, he started a company called Professional Computer Systems (PCS) with friends and worked there for a few months. At PCS, he developed a Nepali keyboard system—similar to today’s Unicode—about 15–20 years before Unicode became popular.
The Pursuit of Knowledge
His thirst for learning hadn’t ended. In 1988, he returned to the UK to pursue a PhD in AI. He completed his PhD in 1993. A year before finishing, he was offered a research fellow position at the University of Edinburgh, where he was studying.
His PhD focused on developing systems that would help computers understand natural language.
In 1994, Suresh realized that it was nearly impossible to fully teach computers human language through coding alone. He believes that language is too complex to be understood purely through rules.
Later, he joined the University of York as a faculty member in the AI Research Group. He worked there for 23 years and eventually became the head of the group. Over this period, he supervised more than 22 PhD students, many of whom now work at top AI companies around the world.
In 2019, Suresh returned to Nepal. Though he had always wanted to contribute to Nepal, it was only in 2019 that the time was right. He is now a co-founder at Wiseyak, a company developing AI-powered health services, including chatbots and call center solutions. Wiseyak has offices in both Nepal and the U.S.
He is also the founder of Numeric Mind, established in 2020, which analyzes data from clinical trials. In addition, he works as an honorary AI Chair Professor at the Madan Bhandari University of Science and Technology.
Suresh is also the Head of AI Research at Fusemachines. In recognition of his international contributions to information technology, he received the Nepali Diaspora ICT Award 2023, presented by Living with ICT.
Creating One of the First Avatar Chatbots
Around the year 2000, when the internet was just starting to spread globally, Suresh was already teaching at a university. In 2000, he and some friends founded a startup called Lexicle.
Lexicle developed a virtual assistant with a human-like avatar for customer service—arguably the world's first avatar-based chatbot, according to Suresh.
This assistant could respond to queries using voice and was implemented in First Direct Bank, a division of HSBC UK. The software took one year to develop and raised half a million dollars in investment.
Lexicle was even featured on BBC Click, a technology program. However, in 2003, the company shut down. Suresh says the startup failed because it was “early to market.”
“Back then, people were just beginning to understand what web browsers and the internet were,” he says. “Chatbots seemed bizarre. We were probably 10 years early.”
(Originally published in Nepali, This is the English translation. Click here for Nepali content. Parts of this text have been translated with the assistance of AI-powered tools.)
