DQ: What is data explosion? How it is changing the current scenario?
PN Sudarshan: The volume and velocity at which data is generated has increased exponentially in the last decade, due to the increasing digitization of business operations, and also the proliferation of digital channels. Now, digital is the preferred channel for consumers to access information, entertainment, and communications. By various estimates, India’s data usage per smartphone, is one of the highest in the world, with an average of 9.8GB per month. It is expected to grow further, with wider network coverage and penetration of cost-effective smartphones.
When we consider the Internet of Things (IoT) ecosystem as well, the volume becomes humongous. By various estimates, the number of cellular IoT connections is expected to be more than 3.5 billion by 2025, which will again generate substantial amounts of machine data. This has a tremendous potential for changing the ways businesses operate, by generating value from the big data, leveraging available tools, technologies, and talent in data sciences and analytics.
DQ: How will data science improve the businesses with regard to future sales, production requirements, market performance, and time-based variables?
PN Sudarshan: Data science is an interdisciplinary field that requires skills across mathematics and algorithm development, technology, and a strong business sense. A data scientist looks at data at the granular level, builds the algorithms to analyze the data, and provides insights for the business problems.
As a discipline, data science provides the framework and mechanisms to generate insights from large volumes of data, in identifying trends, detecting anomalies, and predicting the likelihood of an outcome, which can help in business use cases across marketing, sales, production, etc.
DQ: What are your suggestions for new engineers looking to break into data science?
PN Sudarshan: Having a strong grounding on mathematics, statistics, programming, along with a good business sense, would help engineers build a career in data science. Strong grounding in core skills, along with experience in the tools and technologies across data warehousing, statistical analysis, reporting, etc., could help them in this field. Being inquisitive and having an eye for detail, are some of the other qualities that would help them succeed in the field as well in the long run.
Dataquest: How are you helping your clients with emerging technologies such as AI, ML etc.? What exactly are they asking for?
PN Sudarshan: While we cannot comment about the specific client requirements, at the broad level, AI and ML helps to create a prediction engine for a specific use case or business application that maximizes accuracy of predictions and minimizes errors. This is achieved as algorithms gain experience in a given context (as they get exposed to and more training data), and the feedback loops built in to the system.
ML could be beneficial for any application that involves prediction, such as medical diagnostics, image recognition, autonomous driving, predictive maintenance, voice assistants, etc. In the context of data science and analytics, AI and ML could be an effective augmentation of current tools and human intelligence, to generate insights from large volumes of data.
Dataquest: What does the future of AI look like? What are some upcoming trends in ML?
PN Sudarshan: AI and ML is still in the early stages of large-scale market adoption. They will continue to witness interest from enterprises, as they gradually move from pilot implementations in identified business problems to broader implementations.
Use cases across customer-facing functions, as well as middle-office and back-office functions would witness technology-enabled transformation aided by AI/ML tools. Functions, such as customer experience management, witness the adoption of AI-enabled chat bots as enterprises look to improve their customer engagement.
In the consumer market, voice AI is one area that could witness increasing adoption in the medium term. Considering the fact that majority of the smartphones these days are shipped with built-in voice assistants, voice AI has a potential for widespread market adoption across multiple use cases, from performing basic tasks such as setting alarms, calendars, etc. to home automation systems, to access and control connected appliances, such as lights, thermostats, security cameras, etc. These are some of the key trends that are driving the AI/ML adoption.