Big Data

The Rise of Data-Driven AI and ML: Impact on Business and Industry

CXOToday has engaged in an exclusive interview with A. Kishen Kumar, Group Chief Technology Officer, ASK Group


How do you foresee the role of big data evolving in the next 5 to 10 years, and what opportunities or challenges do you anticipate for businesses in leveraging this technology?

“Data is the new Oil” is the moniker of our times and has always played a key role in all walks of life. Earlier data was merely recorded, and it was a manual task to ensure accuracy and usability. Accuracy at any stage, right from input to its interpretation / results was also largely dependent on how it was handled. Most of the super computers were developed between the 60’s- 80’s, and today they’ve evolved to not just store our data but also process, analyse and share ‘learnings’ with us.

In the present era, I think it will drive the business models and not vice-versa. With a data-driven business model, there are fewer ambiguities and the element of human biases can almost be eliminated. Data also enables higher accuracy with greater speed at a significantly lesser cost.

Data driven business models guarantees a more accurate customer analysis and behavioural patterns.


As a CTO, what strategies would you implement to ensure that your organization stays ahead in utilizing big data effectively to gain a competitive advantage in the industry?

At ASK Group, the source of data, its integrity, and continuous sustenance has been the DNA of the organisation. We have focussed on complete Data Governance as a first step and a key driver is our Data oriented Growth strategy. With the advent of new analytics and algorithms, the focus is on using data to identify and drive new business avenues; focussing on growth and expanding to reach newer geographies. This has reaped good rewards and is paying good dividends. Data Analytics and Science is an ongoing process. The idea is to understand customer behaviour, analyse patterns on investments, identify business strategies for specific segments and target customers, which is a continuous ongoing process. As a small example: real time investing via digital channels was implemented at ASK in August 2019 and then subsequently using Aadhar based authentication and seamless integration with NSDL, CKYC and CDSL using API has been done.

All these have come in handy and the business volumes have increased significantly, giving us an edge over our peers.


With the increasing complexity and volume of data, how would you address concerns about data privacy and security while still harnessing the potential of big data?

Big Data is always overwhelming and there are concerns about maintaining integrity and privacy. With advent of Digital India and APIfication of most businesses and service providers, good quality data can be sourced from regulatory bodies like CVLKRA, NSDL, Adhaar, Pan etc thereby improving Data Quality issues.

The technology has also kept good pace with data volumes and bigdata in the form of data lakes, warehouses etc… Today these are a no brainer when it comes to the storage and efficient processing.

However, when using various algorithms to create various models in the process, it is still a challenge – like in previous years VAR calculations would go on for days but that resolved and behind us. So, I don’t think that should be a challenge with emerging technologies in terms of CHIP and memories and their ability to handle LLM (Large language Models).

With data regulatory bills like the GDPR and DPDP it will be a key factor to ensure that we have customer consent to share and utilise data the way business would like to.

The challenge of acquiring and management of customer consents, registration, de-registration of customers etc will play a pivotal role in future and also needs to be reviewed for any impact based on Data Protection and Privacy bill. This could prove to be a challenge for businesses to interact with customer, since without their consent and understanding, customers might lose access to products that might be suitable for them or miss crucial updates.


Real-time data processing and analysis are becoming more critical in BFSI. How do you plan to integrate real-time big data analytics into your organization’s existing processes and systems?

Indeed, the concept of offline for data and data processing and has been one of the banes in the financial services businesses. Some of them are being driven out of business or technology considerations. The costs associated with real time data also needs to be justifiable from a business perspective.

From an organisations or business point of view, real time updates on customer profile behaviours across channels, merchants and others, the behavioural patterns can be observed and interlaced with real-time experience. With humans, the emotional quotients is a key area and if that pattern can be matched and analysed then it would be result in a bombastic business revolution.

ASK has been a pioneer in the adoption of digital technologies and has won multiple awards for the same. We have the early adoption advantage since the core transformation of the back-office had been done as early as 2018. This has provided a very stable environment in terms of Operations, Financial reporting, Customer reports, and most importantly, business compliance as per SEBI. This ensures sufficient guardrails for our clients.


As we move towards a more data-driven culture, how would you ensure that the decision-makers across different departments understand and trust the insights derived from big data analytics?

I wouldn’t be surprised if soon, business heads or CEOs of most organisations are indeed CDO (Chief Data Officers) or people who are able to understand and analyse data. It will be mandatory for business heads, operation heads, and functional heads to understand how the businesses are built on data models and should be able to quickly adopt and adapt to the changing behavioural patterns of customers. This is the model in some companies today as well.

The advantage of working with data is that, if the input is not tampered with, data analysis will give clear results in a faster and more cost-efficient manner than erstwhile methods. It is also quantifiable and can be tested with multiple what-if scenarios. This also helps reduce several costs such as overheads, storage, but more importantly the cost of errors.


How do you see the convergence of big data, artificial intelligence, and machine learning shaping the future of our organization and the industry as a whole?

I do not see this as a convergence but rather as one of the steps for achieving AI in its true sense. Data capture is the first step; dimensions and analytics, the second; predictive the third, machine learning using NER (Named Entity Recognition), NLP (natural language processing), neural networks and finally amalgamation resulting in AI. These are steps to be followed and I think we are in the third and fourth step initial phase. There are miles to go but unlike in the past where these would take years; for instance, in the 1950’s to the 1970’s when white papers were written on AI and published by various scholars globally; we will soon see the reality in the next 3-5 years. Many FinTech’s and the large IT conglomerates are already on the job. The AI game is ever changing, and I see it making a huge difference to our lives and the way we will adapt to such situations. I think it would be like JARVIS, the popular character from the MCU movie Iron Man, named after the Stark family butler, which evolved into a sophisticated AI assistant that runs the house and is an acronym for Just A Rather Very Intelligent System. And the newly created being with emotional intelligence, Ultron etc.…just to relate figuratively.


Data visualization plays a crucial role in conveying insights effectively. How would you encourage and support the adoption of data visualization tools and practices within your organization?

This is perfect question as the whole thing about Data is its visualisation. How does one visualise data and this depends on how various roles in the organisation would like to perceive, review, and analyse data for their future business. Here we are not talking about AI etc. Let’s take an example:

CEO: Depending on his/her KPI’s:

  • Business Growth: Numbers, Margins, NIMS, NPA, Channel growth, Product growth, Operational efficiencies
  • Product: growth in existing products, new products launch, behavioural patterns, performance of the products, customer centric views
  • Customer: New customers, nos of exits, new branches, top customers, probability of the customers leaving, customer wallet size and potential, creation of new products.
  • Customer Services: Nos of calls, TAT, inefficiencies, unhappy customers link the number of exits to the customer service.
  • Regulatory: Asset Liability Mismatch, Reviews, Basel 3 norms, Capital Adequacy
  • Marketing: campaign and its efficiencies, press releases, interviews, product sales, market commentary, reviews etc.
  • Technology: New avenues, ROI, usage, Security, technology driven business growth, uptimes, challenges etc.
  • HR: Joining, exits, employee behaviour, Satisfaction Scores etc.

Similarly for Sales heads, and operational heads have their own way of relating to the above data. Hence it is mandatory to have the data in the most granular from and follow a bottom-up approach unlike organisations to-down approach to ensure the integrity of data across all streams


With the increasing adoption of cloud technologies, how would you optimize our cloud-based infrastructure to handle big data processing and storage efficiently?

I think adoption of cloud technologies was designed for the term used in the past as SMAC (Social, Mobile Analytics and Cloud) and that says it all. With the adoption of cloud technologies, the pay per use and ability to handle increasing growth volumes is much easier effective and efficient. Optimisation of the data in the cloud is like managing any other cloud centric applications. In terms of processing the same logic applies in case additional thrust or power at critical time the pay per use model for CPU, memory and storage pays off, in elastic mode and the data churn is very important.

In some cases, the costs could also be a prohibiting factor since the volumes are large and processing needs are high, and this needs to be managed well. Few instance, the vintage of data, active customers, inactive customers, and many other factors also help in managing this on the cloud. But it is a no-brainer that cloud, as a platform, is one of the best ways of managing data growth effectively and efficiently.


How would you foster a culture of data-driven decision-making within our organization, encouraging teams to rely on data insights rather than gut feelings or intuition?

 It is very important to foster the culture explaining the importance of data not only within the organisation but also from an external point of view. This needs to be top-down approach of explaining to the end users be it in Customer Support for e.g. to understand customer needs, behaviours, Similarly business needs to understand the importance of managing customer data and derive insights for driving sales, understand, patterns.

In ASK the data driven culture has been driven across the organisation for the last several years and today we are at the cusp of next level of using data for driving new business models. In ASK each business is responsible for its own data as a part of the Data Governance process and is responsible for owning managing and driving its usage.

The business groups have been able to drive and manage its own business plans, interact with customers, and predict many of the futuristic actions. It is preparing for the next steps of machine learning can help in understanding customer patterns, sensitivities and then using AI can suggest better products, optimisation and balancing of portfolio and increase the share of wallet.

It has enabled the customer service team to interact with customers better and follow-up. The next step for them is to get into trained ML models which will help in the complete automation of customer interaction based on the history of interactions and drive the customer satisfaction scores.


What emerging trends or technologies in the big data space do you believe will have the most significant impact on businesses in the near future, and how would you prepare our organization to embrace these changes?

The emerging trends in the data space, as stated in the above replies, will drive organisation business model and strategy from on customer behavioural patterns, investment decisioning, real-time risk management & decisioning. It will have more capability to churn out models achieving high rate of accuracy, better predictability, and reduce risk.

The customers will be spoilt for choice and will be very serious about how and what data they share with partners, businesses, as well as socially.

We as an organisation are in the right direction of having setup data governance, and ensuring data quality is upto standards. The business models encapsulate our way of business strategy, customer and partner driven propensity. It would be an interesting way of life for our business being driven by data in the next couple of years, and run a complete AI based business model yet ensuring our business values of customer first, Integrity, expertise, collaboration, transparency, and efficiency to a sustainable business.



 The interviewee is the Group Chief Technology Officer, ASK Group. The views and opinions expressed here are personal.

Source link

Most Popular

To Top