In this exclusive interview with Express Computer, Vinod Sivarama Krishnan, Chief Digital and Information Officer (CDIO), Essar Capital, delves into the transformative impact of emerging technologies on the industry. He highlights key trends such as real-time data streaming, artificial intelligence, and low-code platforms that are reshaping business processes and driving innovation. Krishnan offers expert perspectives on the evolving role of CDIOs in balancing innovation and problem-solving, integrating AI into business workflows, and the importance of strategic digital transformation. He also explores how organisations can navigate the challenges of data privacy, sustainability goals, and generative AI implementation. Drawing from his extensive experience, Krishnan provides actionable insights on how companies can leverage technology while aligning with broader environmental and societal objectives.
What are the most significant technology trends you see shaping the future of the industry, and how should organisations prepare to leverage these trends?
In the last few years, there have been many exciting technologies and trends. To my mind the most significant technology trends are :
Real-time (streaming) data through IoT: Due to relatively lower cost of sensors and rising telemetry, as well as smart device adoption, we have seen most electrical and mechanical equipment built with smart technology, i.e., the ones capable of building in the premises to collect and store or transmit data on multiple parameters that can easily be streamed. This is because, given relatively low storage costs as well as the development of technologies to handle and store massive amounts of data, it has thus been possible to collect and utilise enormous volumes of streaming data. This data has two uses, one which is at the elemental or instance level, which might be used in troubleshooting or in customer service and deals with one record at a time, and the other which is about identifying and detecting patterns, as we will see in the next technology.
Artificial intelligence (as distinguished from generative AI): At the same time that we now have ready access to good quality streaming data, we are seeing maturity in a technology that has been around for years, but that now end is coming of age as advances in machine learning and deep learning algorithms that can identify complex patterns in big data sets. All such intricate patterns, which arise from complex processes involving various parameters and factors, including all those that may themselves be related, do not appear to offer any obvious solutions using the standard programming and analysis methods but are perfectly suited for deep learning algorithms, which can process the data extremely easily and predict them with great accuracy.
Low Code/No Code Development Platforms: This has been one of the historical challenges for IT teams and leaders over the years in terms of increased workload of change management, especially on legacy systems or even major enterprise platforms like CRM or ERP. This sucks up a lot of IT bandwidth at all levels—from technology up to project and release management—and accounts for a big chunk of IT spending and budgets. Historically, a fear had always existed to not put the functionality in front of users and key users for fear that users wouldn’t be able to understand complex workflows or processes or cross-functional impact and also create some new processes that break existing dependencies and cause business-level disruption. Now Low Code/No Code brings citizen development or citizen application support where users can establish their own processes and start working with data and code in order to drive business outcomes. Of course, there should be good quality controls and code and version management processes, but if controlled well, it can be a massive productivity enhancer for IT teams and simultaneously solve the ills of business involvement and ownership of processes, besides change management and adoption of new processes. Running such hackathons involving large cross-functional teams with practical and implementable solutions is now possible due to such tools.
An organisation should therefore identify key mega initiatives that are strategic in nature, especially those that are customer-focused and revenue-accretive or employee-focused and cost-reductive in nature. Because it is probable that the historical roadblocks to these initiatives from an IT point of view emanate from one or more of the challenges discussed above, they should use one or more of the above technologies to remove the showstoppers and achieve breakthrough innovation in solving these problems and making these initiatives feasible.
How is the CDIO role evolving in this rapidly changing technological landscape, and what are some key skills increasingly becoming important for leaders in that role?
There are multiple trends making it more difficult for IT leaders in today’s world that lead to erosion of confidence and trust in IT leadership’s ability to continue to deliver secure and effective solutions to evolving business needs. Some of these are:
The balance between problem solving and innovation CDIO’s who are trying to move towards innovation often find that while businesses would like to be innovative and have access to the best and latest in technology, they also need robust transactional systems that are capable of handling their increasingly complex requirements without needing to be rebuilt from scratch (since that luxury is often unavailable or unaffordable!). This also puts the skills that are required to solve the day-to-day problems in stark contrast to, and often not compatible with, those required to innovate at an organisational level. CDIOs need to understand this tension and architect themselves and their teams in order to be able to do both, potentially in a more non-traditional way through driving simplicity, increased business involvement, and changing partnership and resourcing models.
Hype around certain technologies: While hype is sometimes an extremely useful phenomenon to secure budgets and create leadership and ground-level excitement, it becomes counterproductive if not aligned with the base strategy or not appropriate for the current maturity of the organisation and its processes. This leads to a situation where business leaders feel their IT teams have not risen to the expectations, and they become mistrustful of their capabilities. CDIOs need to invest in time to understand where and how these technologies can be applied, to get across the facts to their leadership teams and peers, and to ensure separation from hype from reality while staying invested in exploration and proofs of concept.
Ability to simply explain technology and separate out the benefits With the popularity of certain technologies (like GenAI) in mainstream media, there has been a general “dumbing down” of the understanding of technology at all levels of management, from boards to management committees right down to the shop floor or office floor. CDIOs have never faced a tougher challenge in the management of the presentation and the influence of their people skills to influence that understanding and to correct possible misconceptions without appearing defensive or resorting to jargon. Boards should thus be able to appreciate the CDIO’s understanding of technology as well as the company and the business situation and navigate through these impasses to deliver value as well as increase the perceived enterprise value through excellence in harnessing technology.
Identifying the right projects and partners: Perhaps one of the most frustrating tasks for a CDIO is determining which portfolio of projects and initiatives best suits the company, that is, which ones are critical to the company’s future and which have clear and quite obvious payoffs. The budget, bandwidth, and more importantly, the attention of the company’s leadership as well as rank and file impose the constraints. Too many low-payoff, bad business case projects will pull on the team, erode credibility, and thus support and undermine the license to operate IT. Too few poorly run projects will have the same impact. The portfolio should ideally contain some sure bets, exciting learning opportunities, and a few moonshots, and the understanding of the portfolio risk should be well managed to optimise the payback.
CDIOs must become better communicators and evangelists and take a portfolio approach to their initiatives to become more relevant to their organisations.
What are some best practices for inserting AI and ML into the existing business process to drive innovation and efficiency?
AI and ML are not standalone technologies that will somehow drive efficiency or innovation without being inserted into every process of the company. Most companies, B2C or B2B, interact with their customers and suppliers (as well as with employees!) on “moments of truth” meaning a vast number of minor episodes, and the ability to influence those moments and make them effective lies at the heart of efficiency and stakeholder perceptions of how good they are. And accordingly, there are two ways in which AI and ML can be integrated into enterprise business processes.
A. Around the edges: Organisations can employ AI to facilitate transactional processes with information. An example of that could be that while a sales representative punches a customer order into the CRM or ERP system, he or she gets online feedback from an AI model running in the background that, given the outstanding balance and payment behaviour of this customer, perhaps he or she can ask for a higher up-front payment or needs to price more stringently to account for the additional risk on account of poor payment history. This “augmented intelligence” will be directly accretive to the bottom line and drive better controllership as well. It is highly visible and directly effective in improving company performance because it is very close to the front lines.
B. Processing truckloads of transaction data at the back end: With processes such as process mining on large transaction amounts, great insights on business-process efficacy can be obtained. For instance, in one company with an existing “No PO, No Pay” process, analysis of the backend process mining indicated that in 30% of the cases, the PO was created in the system after the invoice date, clearly showing that the process was being beaten. A closer look would find that for that 30%, it was indeed the industry practice to have a more just-in-time approach, and perhaps the right answer would have been Blanket POs or a different process than for the other 70%!
With the growing importance of data privacy and security, what strategies would you advise organisations to use to protect their digital assets while enabling digital transformation?
Given the data privacy and security problems, there are three steps CXOs must take to defend their digital assets.
Understand the critical few elements and the real risk, not the optical risk: The greatest challenges are made around identification of those elements that really matter to the enterprise—not those that are the most visible. Often, visibility of enterprise data externally is a function of the functionalities that an enterprise offers to its customers and partners. Therefore, a very easy way to be sure of security would be not to provide such functionalities at all! That is clearly a false choice. So savvy CDIOs have to balance the risk of not providing those services, which can create benefits to customers and partners, against the incremental security risk that comes with supplying them. This will typically be done in a manner where customer and partner experience happens under some form of strong safeguards to prevent mass exfiltration.
Protect that which can and must be protected: Much like the above point, it will not be possible unless prohibitive, at least from a cost and effort viewpoint, if not the management and team bandwidth required for the same. The fewer key items to protect, the better they can be protected.
Prepare for challenges: Much as organisations work on prevention, they also need to work on processes around eventual breaches and subsequent impacts. At least half the security efforts need to go into the remediation processes around efficient and fair handling of breaches, quick assessment of impact, corrective actions needed, and systemic fixes to these problems.
How can organisations balance the objective of technological innovation with sustainability goals, making sure that their innovations do well for the environment and for society?
Digitisation often plays as a productivity aid; hence, in most cases, it can be straight away correlated with lower manual efforts, lower energy consumption, better energy and material usage, and waste avoidance in the processes of the company. It thus makes sense that if digital technology helps the company to process more transactions using fewer or equal resources, then per transaction basis, it is saving a lot of energy spent and, thereby, making the world a better and more sustainable place. Hence, if the digital technology implementation avoids the need for travel—or for an equivalent physical transaction in general—it then gives way to an outsize positive impact from the sustainability angle.
Growth in energy usage for computation is still only a minute percentage of total power usage, and so it means that a small addition of energy used for digitisation and digital technology like LLMs or data centres will power a disproportionate reduction in energy usage in the real world.
Lastly, it is in the nature of technology to become more energy-efficient over time with improvements and productivity being driven (if nothing else from a profit motive by IT service and product providers). Moore’s Law operates to drive down prices and ramp up the energy efficiency of the unit compute, unit network, or unit storage.
Could you give me specific examples on the deployment of generative AI in your industry, its impact on business operations?
a) Scalable productivity: Sizeable reduction of paperwork and emailing by automating the routine processes that add little value to the company.
b) Support aggregation of a large set of data and produce summaries: Summarising and encapsulating information across a large and diverse set of data, which effort-intensive is not often very complex, ideal work to be automated.
c) An automation that eases the tracking and monitoring processes: Much of the work, especially in compliance and management supervision, was tracking and monitoring for exceptions and following up to get RCAs or explanations for anomalous data points. What’s great here is that the work can be done away with through generative AI automation that could directly reach the anomalous points, allowing resources to spend more time in analysis and solutions rather than data processing.
What in your opinion will be the major obstacles against the general acceptance of new technologies, and how can firms overcome such obstacles?
From the above all, every answer has pointed out roadblocks to success, and it all starts with knowledge and then proceeds to understanding and then to changing and adoption. Emerging technology is inherently challenging to use, and it needs to be appreciated at a Board and company management level that this is a portfolio approach where some will work and some will not work, at least at first when any of those are attempted. Understanding the concept of a paradigm and the need to have patience – and sometimes, faith – will go a long way in giving IT and Business Teams enough time and space to overcome these challenges.
Like all good portfolio managers, CDIOs can de-risk their portfolio through the following steps:
a) Has an IT strategy that takes into account processing new technologies – By design, the IT strategy should be open to the introduction of new technologies and allow for a lifecycle and replacement strategy for all key elements of the strategy. This will allow companies to maintain continuity and not have to rewrite the enterprise IT strategy each time some new technology emerges.
b) Have an aggressive set of Proofs of Concept to understand technologies, their maturity and potential impact as also costs and benefits – It is important for enterprises to visit emerging technologies that mature over time and understand the impact these may have on key processes in the organization. Not all of these POCs will-or indeed should-succeed, but the enterprise will be acutely aware of the pluses and minuses of the technology and can watch out for what is changing that would impact them, especially the shortcomings which usually get addressed within a generation or two of those technologies in their development cycle. If we have a way of monitoring technologies, especially those that failed – understand why they failed and keep track records of what developments could re-trigger interest in these technologies, then we are likely to be quicker to adopt them when they are ready for our reality and realise the potential of those technologies.