In this article, David Quero, Zigurat's professor of Global MBA in Digital Business and EMEA Program Excellence Leader & Delivery Manager at IBM Watson (IBM Artificial Intelligence platform), dissects the role that predictive analysis plays in digital transformation in general and in the advancement of artificial intelligence in particular.
In general, Digital Transformation is something that is not as much talked about as it was in the past, but it’s something that is already happening.
When we take a look at small companies, typically the least innovative segment, we find they are already transforming their processes by implementing:
- 24 x 7 support based on virtual agents
- Salesforces empowered with mobiles and/or tablets that enable them to provide cost estimations to prospects and even to close deals
- Use of geo-location apps to optimize routes,
integrated IoT in key components to avoid extreme accidents,
This is not something from a sci-fi novel, but a new reality.In the area of Predictive Analysis, we see how exploiting data is not only a nice-to-have gadget anymore but has become a must. Even the simplest businesses need the advantage that data offers to foresee the demand and how to offer the right products to the right clients. Data has become the key to staying competitive and surviving.
When we analyze the principal elements that fuel the digital transformation, we find that cloud adoption keeps on progressing. Even the most data-security-concerned industries (health and especially financial services) are moving part of their business to the cloud and accepting that they need to manage the risks. The option of avoiding them completely by avoiding the cloud does not exist anymore.
Disciplines related to Artificial Intelligence (Visual Recognition, Natural Language Processing, Audio Recognition – also known as Speech-to-text, Machine Learning, etc.) stay on their track:
- On the one hand, the constraints are better understood and the solutions created are more realistic, which can be seen as some disappointment or at least expectation decrease from the market side (the post-hype phase in Gartner curve).
- On the other hand, some fields keep developing more rapidly. Speech-to-text has improved remarkably in the last years and Deep Learning, a type of machine learning, is opening new doors to explore.
At the same time, those techniques are becoming more and more relevant in combination with predictive analysis. As we have implied, we cannot expect that natural language classification creates the perfect virtual agent (chatbot) that is able to perfectly understand any question and keep any conversation. We have realized that the value of classifying natural language and converting it into a variable that can be managed by a predictive model is enormous. And the same about visual recognition, audio, etc.
Now, instead of searching the absolute disruption with pure artificial-intelligence cases, most companies are taking advantage of the fact that classical predictive models can be fed not only with structured information but also with unstructured information (80% of existing information) that can be pre-processed with AI algorithms.
This humble/silent use is, in fact, much more powerful as it’s changing most of the processes in the services and goods that any of us consume every day, making our lives easier and providing us the advantage of more efficient services, good creation and distribution.
Take a step further with predictive analysis today!
Professor of the Global MBA in Digital Business
Program Excellence Leader & Delivery Manager at IBM Watson