Beauty of deep learning lies in ease of implementation: AI expert Dr. Kolluru

One of India’s biggest names in AI, Dr. Dakshinamurthy V Kolluru, took to the stage at Rakuten CTO Summit 2018 on March 14 in Bengaluru, also known as Bangalore.

Speaking in front of a rapt audience of 51 CTOs and heads of engineering from Rakuten Group’s businesses across the globe, Kolluru traced the fascinating history of machine learning, deep learning and artificial intelligence, elucidating on how AI can benefit businesses and improve customer experience.

Rakuten, which promotes use of AI in all of its group companies, was keen to hear the thoughts of the man described by Analytics India Mag as “a visionary, an analytics expert and a passionate educator, who has been doing highly innovative work in the field of analytics — be it consulting, product development, corporate training or educating — since 1999, when analytics was not part of the common lingo that it has become today.”

The Founder and President of International School of Engineering (INSOFE) Hyderabad, Kolluru has helped set up many data science centers of excellence (COEs) and has conducted training for multinational corporations such as Johnson and Johnson in the US and Microsoft, HP, Broadridge Financial Services and others in India.

Kolluru, whose expertise lies in simplifying complex ideas and communicating them clearly, drew on landmark studies to explain where AI and deep learning fit in the spectrum of technologies like machine learning and robotic process automation (RPA) and how they can help complex businesses like Rakuten solve problems across functions and verticals. “The beauty of deep learning is not the difficulty of its implementation, it’s the ease of implementation. What really excites me is when an undergrad student tells me ‘Look, Dr Murthy, I just applied deep learning and beat the benchmark you created after six years of your machine learning work,’” said Murthy.

Quoting AI pioneer Andrew NG, who in an HBR article wrote: “If a typical person can do a mental task with less than one second of thought, we can probably automate it using AI either now or in the near future,” Kolluru said that tasks like identifying an object, understanding a text, reacting to a signal — all of which a human can do in less than a second — can be done by AI today at human level accuracy.

Kolluru has been working closely with Rakuten India for the last six months to implement deep learning in various products and solutions being developed by the India team for the group’s global e-commerce and internet services. At the summit, he conducted an exercise for the Rakuten CTOs that required them to think about the business areas in which they can most effectively deploy automation. “When you are embarking on this journey, you want to identify those business problems that have enormous ROI and are easy to solve,” he said, nudging them towards the deep thinking process required to implement deep learning. He also asked them to identify the challenges they foresee in implementing deep learning and assess which delivery models would suit their business needs.

At Rakuten, the AI Promotion Department accelerates the commercial deployment of AI functionalities to various Rakuten services. This can be seen in Rakuten’s Ichiba marketplace, which is already leveraging AI to predict product sales, classify products, analyze user ratings and reviews, and improve recommendations and search for customers, as well as in Ichiba’s recently introduced FAQ-based chatbot for 24-7 customer support. AI functionalities have also been utilized by group businesses such as Rakuten Travel, the Tohoku Rakuten Golden Eagles baseball team and others.

On C2C e-commerce platform Rakuten PriceMinister, developed in collaboration with Rakuten Institute of Technology (RIT) Paris, deep learning algorithms are used for the purpose of image recognition. Rakuten Viki’s innovative Learn Mode feature, developed in collaboration with RIT Singapore, deftly combines language learning with television watching and also uses artificial intelligence and natural language processing.

If you ask the CTOs and engineers who attended Kolluru’s talk, these examples are just the tip of the iceberg when it comes to the potential AI holds for the Rakuten ecosystem and its members.

Read more reports on CTO Summit 2018 here.