Cloud and Cognitive Services

In May 2017, Prime Minister Narendra Modi said Artificial Intelligence (AI) would drive human race. It could be debated whether there would be any job left or not. However, experts say there is huge possibility of job creation through AI. It has generated immense curiosity and the developments, which India as a nation would be undergoing in the coming years as part of the Digital India programme. Before delving into AI, it is important to understand its meaning. We should not get swayed by the various associated terminologies like “Machine Learning” and “Cognitive Learning”, which attribute to cumulative digital development of the nation. This also denotes the untiring efforts to keep up the pace of the technological progress of the developed nations like the US, which has recognised India’s potential and in turn leveraging India’s strength in developing new technologies and tools for better services based on the needs of the customer.

At present, in India, AI is being used in various sectors such as business, banking and insurance, financial and electronic transfers and health. In business, BSE Ltd (Bombay Stock Exchange) has implemented a project called “analytics rumour” successfully using AI. It involves surveillance of the print and social media like Facebook to capture all information and building up intelligence to determine if the information is an existing rumour or something new or fake. Moreover, AI is used to correlate the movement in the stock prices. Earlier, this work was carried out manually through a department, which had 10 to 15 people scrolling all the newspapers, website etc, to identify and collate information manually. This collated information was recorded and kept as base, which the management referred in times of any enquiry. But thanks to AI, the time consuming work of filtering and collating information is now fully automated. Supervisor now just needs to validate the filtered news. This has led to increased efficiency and saving of a considerable amount of time. Kersi Tavadia, Chief Information Officer, BSE Ltd, says “Technology is going to be the future of how things are going to be in India and nothing will be successful unless you build it to run for scale”.

In the banking sector, AI has helped in taking prudent decisions and more crucially has focused on the “edge”, where there is a lot of computing happening at the branch level. The traditional way of just equipping at the centralised level has given way to various activities at the edge due to the present trend of “self service model” of the ‘equipped individuals’ who have phones, which are getting smarter, bandwidth getting stronger and propels them to do a lot of computing themselves for ‘instant gratification’ or ‘instant requirements’. As Rajesh Borde, Country Manager, BFSI, Hewlett Packard Enterprise, points out “edge is your place where you require a lot of data churning as there is a lot of analytics going on”. In addition, AI is extensively used in finding out the Non-Performing Assets or in simple words ‘referring to loans that are in jeopardy of default’. Complex machine learning and social media in AI help in recovery of loans and is also used in predicting defaults.

Shiv Kumar Basin, CTO, State Bank of India, says “In SBI, AI has been very helpful in promoting student loans and in developing natural language processing and cognitive engines.” SBI has been using AI in student loans as it enables to trace and establish contact with the students living aboard and are in good jobs for successful loan recoveries. SBI, in collaboration with IBM, is developing a cognitive engine in the lines of IBM’s WATSON, which deploys a “chat bot” or online financial advisor, which can predict the top-50 questions asked by the customer by making it simple and by using different kinds of grammatical sentences in various scenarios which are customer friendly. The chat bot also helps in staff education through as it assists the front office application for banking staff, which can respond to customer questions faster. Moreover, chat bot are provided in customer service helpdesk to similarly respond to customer queries and complaints. The drawbacks with these cognitive engines are that it is hosted on the Cloud, which is US based, while the RBI guidelines mandate that none of the data should go outside India. To counter this problem SBI has over the period of three years developed its own indigenous private Cloud called “Megdoot”, which can save cost and facilitate real-time authorisation for storage. Keeping in mind the culture of instant requirement, AI is used in facilitating the scientific risk assessment and credit rating at the branch level. For instance, today a customer cannot be asked to come later for a loan and to decide on an EMI. The risk assessment and the credit worthiness of individuals is being decided in hours by doing analytics at the backend as there are linkages of data of an individual related to his credit worthiness.

One of the most transformative things which is related to AI, is the movement towards electronic payments. A very large number of people have moved into e-monetisation after demonetisation. This throws up some newer challenges in terms of disaster kind of a situation in ‘Cloud’ and importance of cyber security. It is to be noted that there is a tendency to back up everything that creates humungous data problem. Therefore, it becomes important that Cloud usage needs to be measured. Unless metered well, it will become difficult to know what will cost you by putting everything on Cloud and in turn, it promotes only the necessary data to be stored on the Cloud. Moreover, features like what all applications you can have on the Cloud, what is the kind of response time given and how your data is secured are all vital criteria, which determine the data recovery in the event of a disaster.

Cyber security becomes pertinent in AI, because of the intrusion that can harm your data. Cyber security can be addressed at the following two levels. Firstly, at the individual level, where one can practice cyber hygiene by not lodging unwanted data on Cloud or share security protocols. Secondly, at the institution side, educate people to use technology in a secured way. For instance, dissuading people to share their passwords, security related information and making it simple. At present, any financial transaction in India involves two factor authentication in addition to a password. The first, card based authentication and second is the OTP either SMS or email based. In the first, data is encrypted in the card. Both of these methods use AI. Another form, which is prevalent and used in Aadhaar system is the biometrics encrypted in your card, which consists of scan of the fingers and retina, which is considered to be safest authentication method in India. Samsung is also working on retina based “Iris Scan” to make its phones more secured. It is to be noted, that few years back Apple security system was breached in US when the fingerprints were lifted from a door knob, which made them to go for finger vein based authentication.

After meeting with Prime Minister Modi in 2016 and 2017, Ginni Rometty, IBM President, stated, “As a country, India has been digitising and also digitising the world. The foundation is in place. Digital is the foundation. India’s next path is to now make the world cognitive. I think, no other country has a more important role to play than India.” These words are not intended to flatter Modi, but India in the last two years has actually witnessed rapid developments in the field of AI. For example, a Bengaluru-based startup KlickDoc has produced an innovative Cloudbased healthcare platform that connects patients with doctors, pharmacies and diagnostic centres with the facility to store and access the medical records on the Cloud. This is somewhat similar to IBM’s WATSON, which plans to collaborate with India in the field on healthcare especially in the cancer treatment. There is a huge scarcity of cancer specialist doctors in the country – only 1 oncologist for 1,600 cancer patients. In addition, Indian artificial intelligence Startups, like the Bengalurubased ‘Mihup Communications Pvt Ltd,’ and Chennai-based ‘Mad Street Den Systems Pvt Ltd,’ are among three early-stage local ventures that are on the list of some of the most low-key, yet promising Startups around the world.

“We cannot write off human intelligence because you cannot take all the advice given by the machines. Since most of the time it will be history based, it becomes imperative for a human analyst to correlate to the present scenario and take informed decisions at a particular level,” adds Borde. This eases the pressure as India does not envisage a trained humanoid replacing human labour, which involves matters of intelligence and discretion based on observation of human emotions.