Balancing Act for Data Science and AI

With ever increasing business challenge, need for faster success and quicker time to market for realising benefit and ROI; digital transformation is at it’s forefront. Data Science and AI is clearly in everyone’s mind and priority at the moment and will continue to drive momentum. At the same time, practicality of situation is getting realised and balancing act is needed as always in the game!!

Some of the key dimensions are as follows. It is important to strike a balance by considering most of these factors.

• Data and Insights - If we follow various infographics and statistics about how much data is getting generated every second, the numbers are staggering day by day. Some of the sources expect that by 2020, for every person on the earth, approximately 2MB of data will be created every second. We keep noticing that 90% of the data was generated over last couple of years. While it is obvious that the important aspect is to find insights from the data which is relevant, it remains an interesting aspect and will continue to require various KPIs for successfully realising value of those data. It could fall into an 80-20 ratio where 80% of cases firms are not sure which data to refer to and analyse on. This obviously put them into a position to struggle and find value out of it.
• Clarity and Explainability - While amazing solutions, models are being developed to solve problems, focus does exist to create an ambience of “clarity” all the time. Therefore “explainability” of models / solutions are getting into a “detailing” stage and this is helping consumers understand “what is happening”, “how it is happening”, "which parameters are causing the target to behave like the way it will behave" and so on. It is not a “black box” anymore. Though it is not that simple, crucial aspect in my personal opinion is that “detailing” is helping answer many questions and thus avoiding confusion and will continue to “improve”.
• Safety, Trust, Security and Governance - This is obviously one of the top priority for firms and with CXO community that they can not take their eyes off. Few weeks back this year, EU released guidelines for ethical use of AI. They do emphasise on the "trustworthy AI" aspect which caters around points such as lawful, ethical and robust. While AI systems can empower and augment human beings in enabling them to make informed decisions and fostering fundamental rights, appropriate oversight mechanisms need to be ensured as well. The OECD (Organisation for Economic Co-operation and Development) has also announced a set of AI principles recently which is based on their objectives. This establishes some key principles such as AI for sustainable development and well-being, AI systems can be designed in a manner that respects rule of law, there should be transparency and responsible disclosure around AI systems to make sure that humans understand those outcomes and can challenge them.
• Usability and Relevance - Customer or end user experience has been the focus always and it is increasingly getting more and more attention by firms / agencies / enterprises who want to stay customer focused and be successful in their business. Hence usability and relevance by use cases, by industries, by potential fitment to the need at ground level is getting refined. If we take retail industry for example, key themes are getting more and more prominent such as - user convenience being "prime" focal, healthy living possibilities leading to healthy retail experience and so on.
• Methodology, Framework, Architecture - The methodology, structured approach, framework, architecture will continue to dominate the way solutions are being defined, designed and developed. When I look back to execution of some of my past Data Science programs which involved end to end execution, I think some way or the other methodology such as CRISP-DM, related approaches, principles and the steps taken look so eminent in defining the success path to the whole program. This is very relevant for Data and AI and play vital role. As we progress forward, this will not change much. The design thinking approaches, agile principles do impose some ways of driving Data Science and AI programs in it's own way.
• Adoption, Acceleration to Growth - Most firms are adopting technology and trying to see how innovation can drive their businesses with the help of technology to be able to build intensity, velocity and continue with the momentum. Adoption is something that not only required to be driven from top or follow a top down approach, it also needs to be appreciated by entire team that are involved in driving business themes and problems. Strategy will be to call enterprises and agencies to ramp up investments in AI research and explore different ways to advance the technology across society.

To sum it up, a balance of all these elements and an effective strategy around them is needed to define success factors and build growth momentum in Data and AI. Each and every aspect is critical considering a particular context and can be strategised accordingly.

Disclaimer: "The postings on this site are my personal point of views from my experiences, thoughts, readings from various sources and don't necessarily represent any firm's positions, strategies or opinions.”