Indian software minds are all set to power a new technological change in the world where software will make decisions for CEOs and leadership teams.
Ray Kurzweil, American computer scientist, author and well-known futurist has made an incredible number of predictions on Artificial Intelligence (AI) since 1990, notable among them being, a computer would defeat a world chess champion by 1998, that people would be able to talk to their computers through commands, and more. More recently, he predicted that AI will reach human levels in a decade and in 30 years it will have multiplied the intelligence a billion-fold.
Every company from Google to Facebook to Oracle to Microsoft to SAP are working on projects and platforms where software can learn and make decisions without having drones of people managing business.
AI is not really a buzzword. Before knowing what impact it makes, we need to know the difference between Machine Learning, Deep Learning and AI. Machine learning is an approach where you teach the machine to interpret the data, while deep learning is an approach where the software learns from data patterns and their interpretations. AI is when the software makes decisions for itself. “These are systems that are going to reshape society in a decade. It will make enterprises efficient and accountable,” says Amar Chokhawala, founder of Reflektion, which uses AI to convert visitors in a website into potential customers.
According to Gartner Inc, market hype and growing interest in artificial intelligence (AI) are pushing established software vendors (like Infosys and Wipro) to introduce AI into their product strategy, creating considerable confusion in the process. Analysts predict that by 2020, AI technologies will be virtually pervasive in almost every new software product and service. Gartner predicts that by 2020, AI will be a top five investment priority for more than 30 percent of CIOs.
“As AI accelerates up the hype cycle, many software providers are looking to stake their claim in the biggest gold rush in recent years,” says Jim Hare, Research Vice President at Gartner. He says that most vendors, unfortunately, are focused on the goal of simply building and marketing an AI-based product rather than first identifying needs, potential uses and the business value to customers.”
AI refers to systems that change behaviours without being explicitly programmed, based on data collected, usage analysis and other observations. While there is a widely held fear that AI will replace humans, the reality is that today’s AI and machine learning technologies can greatly augment human capabilities. Machines can actually do some things better and faster than humans, once trained; the combination of machines and humans can accomplish more together than separately.
So let us take a look at industries that will see rapid changes in industry verticals and horizontals:
Banking and finance
Banks will look at their IT vendors to automate loan assessment and risks. Beyond automation of information, AI platforms will be able to look at credit default patterns of customers and suggest action on increasing or decreasing interest rates. The pattern recognition and actionable insights provided by the software can reduce Gross NPAs. “AI uses different data sources and recognises patterns in loan repayments. It can also use a lot of information to assess a company or individual’s credit worthiness,” says Ashwini Anand, founder of Monsoon CreditTech, who is building AI to assess risk.
Companies like NextAngles, based in New York have at least 30 data scientists creating an AI platform that can track fraud in the stock market.
An IDC report says financial services IT spending will reach almost $480 billion worldwide in 2016, with a five-year compound annual growth rate (CAGR) of 4.2 percent. This money is used by the financial services industry to trace every possible route of money. Yet, 95 percent of the time money launderers outbeat the system by opening multiple fictitious accounts and companies across the world. Therefore, NextAngles is building an expert system to tackle this persistent issue.
“The core intelligence comes from knowledge models which are deep domain models on particular aspects of banking work. This is different from conventional big data-driven data science. The core aspect of such expert systems is a branch of science called ‘ontology’,” says Mallinath Sengupta, co-founder of Next Angles.
Ontology in computer science looks at the relationship of that which is (money), the properties that interact with it (banks) and the events that change the relationship (money laundering activity).
According to PWC, banks are asking IT vendors to do the following.
- Update IT operating model to get ready for the new normal, which is AI and Blockchain.
- Slash costs by simplifying legacy systems, taking SaaS beyond the cloud, and adopting robotics/AI rapidly.
- Build technology capabilities to get intelligent about your customers’ needs.
- Prepare the architecture to connect to anything, anywhere.
- Pay enough attention to cyber-security.
- Make sure there is access to the talent and skills necessary to execute and win with new IT projects.
If this was in the realm of finance, the customer service industry is changing too.
Today customers are already used to a robotic voice when they contact a BPO. But, then the problem here is that customers have not been identified by the years spent using the service. Each transaction is isolated and a loyal customer is upset over the treatment meted out to him or her while getting a query answered. AI can solve this problem of comparing transactional history across a few years; it can even draw data from social sources and then lead the customer to a robot agent most suitable to solve the problem.
“Major corporations outsource their support and yet they are not able to manage customer churn. AI can solve it for them,” says Vasudev Bhandarkar, CEO of ScoreData, who is building such a platform and working with 17 financial customers in India and globally. He says AI can pinpoint an important customer and sieve through the noise. Any AI platform can identify patterns from millions of accounts – on a real-time basis – and tell the organisation why a particular customer is important. It looks at their behaviour based on the billing cycle and the type of queries raised over a period of years.
Recently, Volkswagen AG announced that in a decade it will place more emphasis on software and will completely change the way it manufactures cars. Christian Senger, Head of the Volkswagen’s e-mobility division says that cars can longer be looked at as a transportation unit.
According to him, AI can come in two forms, in one it can be used to serve consumer apps that can become the interface between the car and real world. Apart from throwing in options of proximity to retail centres and promotions, the car’s AI bot can decide, based on your consumption habits, whether to platoon itself towards the parkway of the retailer or just order the product on its behalf.
“While this is the futuristic element that every car company is working towards, it is a journey that we have to take by collaborating with tech companies, which includes startups,” says Senger.
However, AI can fit into the entire value chain of automobiles, right from manufacturing to retailing. Big Data concepts have already been used by dealers and insurance companies to prescribe and predict component failure and driving patterns.
An AI platform can help reference millions of customers into buckets of data and help companies predict long-term value (like propensity to return to the brand) of these car owners. So data is essential to an AI platform and its best use is in finding patterns in consumer behaviour.
Have you ever wondered what a CEO goes through in a consumer company? He has to handle thousands of markets and stores. Added to that is the clamour of e-commerce and the entire supply chain that needs to be handled. Tucked away in a small building on Lavelle Road, in Bengaluru, is the $100 million Manthan Analytics, which is building a voice-based AI platform called Maya. With Maya, a CEO can have access to information from a given market at any time and figure out how sales were and why they dropped on each day. The system tracks social feeds and correlates it to market performance.
For example, if social feeds give you information that there is major roadwork on the weekends, then it can also pinpoint why people have not walked into the store on a particular day. Similar work was done by Walmart where they could plan the supply chain based on weather data. So during a hurricane, Walmart would have all the essentials available to city dwellers rather than stocking up on inventory that would erode Walmart’s margins by not selling.
“Clearly AI platforms can pull data from several sources like Enterprise Resource Planning (ERP), Customer Relationship Management (CRM), and social data before citing the reasons for the uplift or drop in sales at a particular store,” says Atul Jalan, founder of Manthan Systems.
Some of these platforms are in the Machine Learning and Deep Learning stage. Let us not kids ourselves yet because AI is all about the technology taking decisions. In the retail industry, it could mean that AI will be able to send a CEO a report on what should be done the next week to improve sales.
“Machine Learning is going to make all engineers focus on larger solutions rather than just patching databases,” says Larry Ellison, Chairman of Oracle, at the Oracle OpenWorld Conference held in San Francisco recently.
Manufacturing and agriculture
Today robotics in manufacturing is not an afterthought, they are pervasive in every plant. While patching robots and understanding their rate of failure has always been measured, the use of AI can help organisations enable the software to manage the plant.
The AI software can reduce loads on one robot and increase loads on others based on production cycle and market demand.
“Bosch has elements of tracking data through sensors and predicting the rate of failure of components in tractors,” says Soumitra Bhattacharya, MD of Bosch India. He says that these are rule-based engines, but in the future sensors can patch the problem themselves.
Even GE Aviation is using techniques such as Machine Learning and Deep Learning to track when its jet engines can wear out, and can, therefore, help their clients save on paying top dollar for servicing a worn out part. “We are already processing huge amounts of data on a real-time basis. GE is making machines smarter,” says Alok Nanda, General Manager, GE Aviation Engineering – India.
AI can become the link to create a common platform to track healthcare insurance, treatment history, and drug sales data. This can become the panacea for the disparate information systems that work in healthcare currently and bring transparency in the industry. But this is easier said than done because of the various vested interests within the care ecosystem who would not like data capture to create transparency. However, AI can play a role in understanding treatment outcomes.
Horizontals are key
For AI to happen, what helped were the fall in storage costs and the rise in computing power. Cloud, data analytics, smartphones, faster networks, and digital security are key components of AI too.
Companies like Reflektion and Euler Systems are trying to define how organisations sample their data better so that the AI software that they use can make better decisions for them.
Consulting firm Zinnov estimates that enterprises worldwide will spend over $470 billion on digital in 2017 and digital spending is likely to increase at a CAGR of over 20 percent to reach $1.1-$1.2 trillion by 2022.
Enterprises in banking, financial services and insurance, retail, manufacturing, hi-tech and telecom sectors will continue to be the top spenders jointly contributing over 60 percent of the total global digital spending.
Today, over 5.5 million people are employed in jobs creates as a result of digital investments of enterprises. This should set aside the worry the AI will remove jobs.
Zinnov’s analysts go on to add that startup investments in areas such as IOT, AI, Blockchain, and Robotic Automation have increased significantly in the last 24 months, with an estimated $25 billion worth of funding across these technologies. This is in addition to over $30 billion invested in acquisitions focused on AI and IOT competencies.
Praveen Bhadada, Partner and Practice Head, Enterprise Digital Transformation at Zinnov says, “Digital is coming out of being a hype to becoming a real spend area in 2017.”
True digital enterprises are investing collectively in digital technologies, digital business models, digital infrastructure, and digital stakeholders as they plan their strategy. New age digital technologies such as Artificial Intelligence, Internet of Things, Robotic Automation, VR/AR, blockchain, drones, and 3D-Printing which saw a minimum uptake in the year 2010, have started proliferating across industry verticals today. But, everything is software. As Vishal Sikka, ex-CEO of Infosys said, “It is software that will make money, not services.”
AI is here and Indian companies have a big stake in the game. India will be the tech stack for the world by building AI services for the world.