Authored by Manish Gupta, Sr. Director and General Manager, Infrastructure Solutions Group, Dell EMC
In the last few years, we have seen an unprecedented data detonation across global enterprises. Businesses have grown increasingly aware of the fact that hard data is the panacea. Data is helping enterprises make faster and more accurate decisions, by eliminating the biases and knowledge-gaps that have plagued enterprise decision-making for decades. This understanding and awareness has rapidly translated into major global corporations across industries giving huge importance and allocating massive investments for infrastructure that can help capture, store and leverage data resources. In the business landscape of today, we are seeing more and more processes getting systematized and robotized, environments and equipment getting more sensorized – with digital IT at the heart of it all. This quantum growth in systematization and sensorization has led to a data-deluge, so to speak, where every minute aspect of business, every process and every decision is reduced to a data point, enabling business stakeholders to make more informed choices in an increasingly VUCA world. Augmenting Business Intelligence with Artificial Intelligence
There is no doubt that traditional Business Intelligence (BI) software has served us well so far. BI helped us understand our industry, our business, our decisions – their trade-offs and outcomes – for several years now. BI helped us capture, organize, store and visualize our data – enabling us to make much improved choices to address our strategic and operational challenges. However, with the data detonation seeing today – far more data, flowing at a much faster rate – will BI suffice in the modern era?
Both common sense and the evolution seen in the data and analytics landscape suggest that stand alone BI is no longer the answer to the complex questions we face in a more uncertain world of business. To drive continued business excellence and competitive advantage in your industry, it is imperative that you move to paradigm where it is not enough to only leverage the huge data repositories available, but also exploit the full potential of data.
This is where Artificial Intelligence (AI) is beginning to get embedded in the organizations and gradually evolving into math houses, powering better data-driven decisions that can help improve efficiencies, for years now. AI can help businesses exploit the full potential of their data. Beyond simply informing and prescribing decisions, AI can leverage the vast amounts of data available to us today and even train itself to uncover insights that can power better choices, leading to superlative business outcomes.
More and more enterprises are expecting their BI tools to bring in some level of self-learning capabilities – where data itself can help create and refine sophisticated algorithmic decision-making systems. Large global tech corporations understand this. To that end, we are seeing increasing evidences for the fact that AI is disrupting BI itself. In recent months, we have witnessed Salesforce introducing Einstein – an AI powered system that automates the analysis of data to uncover key trends in an organization’s marketing and sales process. Microsoft too is putting a huge thrust on AI – their incumbent Power BI tool incorporates various flavours of AI to discover actionable insights in data.
From Descriptive to Prescriptive Analytics
We have seen a dizzyingly fast revolution in the data landscape over the past few years. To understand where we are headed a bit better, it is imperative that we first dwell a bit on how we got here. By mapping the historical journey of data and analytics paradigms, we will be able to better decipher what’s in store for the future.
Descriptive Analytics
The early paradigm in data and analytics – which lasted several years – was what we now call descriptive analytics. Descriptive analytics, in its time, was a massive game-changer in itself – helping organizations understand and interpret data from historical events. Simply put, descriptive analytics served as a sort of ‘post-mortem’ report to help stakeholders answer the question - ‘what happened?’
Predictive Analytics
The big challenge with descriptive analytics – which is obvious to us now with the gift of hindsight – is that it didn’t help clear the fog on what could happen in the future. It seems almost primitive now to rely on data to solely analyse events that have already happened. That said, we did see some traces of past data being used to predict future outcomes in the era when descriptive analytics was in vogue. With the advent of predictive analytics and the rapid strides made in it in the past few years, we were able to systematize and create an entire science around prediction and forecasting. The key thrust of predictive analytics was to answer the question - ‘what could happen?’ This was a crucial step in the evolution in the data and analytics landscape as it empowered business leaders to forecast the future with great clarity and accuracy, in order to make decisions that augur well with their interpretation of the future.
Prescriptive Analytics
In Prescriptive Analytics, we see the early shoots of AI in decision-making. From simply forecasting the future, prescriptive analytics enabled users to make more informed choices from the data available to them. Prescriptive analytics evolved to answer the question - ‘what should we do?’ - by employing advanced statistical methods that only predicted the future, but also suggested multiple courses of action along with their possible impact. With prescriptive analytics we were able to institutionalize the process with which we quantify the impact of various decision options and identify the best course of action. In effect, prescriptive analytics helped business stakeholders get even closer to the holy grail of unbiased, fully data-driven decision-making – by not only showcasing options, but also providing clear and unambiguous recommendations.
The Next Frontier in Data-Driven Decision-Making
So what is next? Where do we go from here? With the advent of AI in decision-making and increasing disruption in the previously staid BI landscape, what are some of the key trends we will see in the movement from BI to AI?
Explosion Real-Time Decision Systems
The next frontier in data enabled decision-making after prescriptive analytics will be AI-powered real-time decision systems – systems capable of taking low-level decisions rapidly as soon as the supporting data is furnished. We are already seeing the proliferation of narrow implementation of such systems across very niche use cases. The ultimate goal would be to replicate the same low-latency and real-time decision-making capability across AI that spans multiple data sets, across use cases to empower cross-functional decision-making. Systems that can leverage this data deluge will leverage cutting edge data engineering and ML methods to power faster and better enterprise decisions.
Opportunities for Data and IP Monetization
As organizations aggregate huge data assets and build in-house intellectual property capable of analyzing and exploiting these data resources, they can harness huge, emerging business opportunities. With increasing pressure on global CIO organizations to transform from cost centres to profit centres, AI-centric systems can be a game-changer for multinational organizations. CIO organizations can monetize and commercialize these specialized data sets and IPS to other organizations and build sustainable revenue streams for their parent organizations. This will be a winner takes all game, as the first past the post to build industry- and function-specific solutions will reap massive benefits in the long term
Skill and Talent Development
Despite huge interest in and proliferation of AI and data science coursework, we are still seeing a massive shortfall in terms of talent available to spark the AI revolution. We desperately need to bring in specialized talent that can help build sophisticated algorithms that can make sense of this data detonation and evolved software tools that are both easy to use and have a robust engineering backbone. Organizations that will win the race for key talent in AI and Engineering will achieve a crucial competitive advantage for years to come.
The AI disruption in BI is here to stay and should be embraced by corporations. AI can not only empower faster business decisions by making sense of the data detonation, but also democratize data better within enterprises and help business users make better enterprise decisions. By bringing data and BI out of silos and accessible to non-technical staff, enterprises can make better, real-time decisions to achieve sustainable, long-term competitive advantage.
No comments :
Post a Comment