Ultra short time to market with AI? It is possible!

Share this post:

Time to market is a distinctive success factor in the digital age. This certainly applies to the development of new or improved products and services, and creating an optimal relationship with the customer. An intelligent handling of data can help. This does not have to take weeks or months anymore.

The role of data scientists in addressing business issues is increasing. But without quick access to relevant data and adequate tools, these professionals can not do anything. However, organizations can not afford to let these highly sought-after employees search for weeks or months for the right data. Yet the average data scientist is involved in this 80 percent of his time. Certainly at a time when companies have to nurture their data scientists, business and IT leaders should offer them the best tools. Tools that are not only available for the biggest enterprises, but could also make artificial intelligence (AI) available to smaller organizations. So they can also shorten their time to market.

Predicting churn

A known application of AI is predicting the churn: does a customer remain or not? This is a simple starting point for many organizations, because of its binary nature. This form of machine learning can for example be hung behind a chatbot. If the technology sees an increased risk of a customer dropping out, the system can manage internally that a real person takes over the contact. Another application example is anomaly detection, or the discovery of deviations. You can also gain insight into customer preferences via machine learning, or offering related products (cross & upselling). Another phenomenon is a textual analysis, such as matching vacancies with CV resumes. A platform for data scientists to use, is Watson Studio: a platform that works for them, without the need for help from the IT department.

Fast deployment

Watson Studio offers a super-fast deployment, open, end-to-end support for the data scientist community. The platform provides for successive process steps such as creating a project, selecting and reviewing relevant data sources, refining, enriching and combining the selected data if desired, and then feeding this to the chosen machine learning algorithm. Everything based on standard integrated tools and features. Ideally, when the data is completely in order, all of this can be brought to a working model in fifteen minutes. Not too long ago people were busy with it for months.

On the one hand, the data scientist must have as much access as possible to data and tools, on the other hand, of course also governance plays a role. In the context of the GPDR discussion, Watson Studio is prepared that certain data can only be shared with the authorized people. In the data catalog these two aspects can be brought into line with each other. The data is available, but always under control.

Low-threshold

The technical details of all this are in fact irrelevant. It’s all about the different steps, which can also be easily understood and followed by the average user. Searching for data, disturbing, refining and modeling – that is all! The essence of Watson Studio is that working with AI on the basis of the right tools can be very low-threshold. This is not a false statement, this is reality!

Would you like to know more how Watson Studio could help your organization? Find out more here. Or are you curious to try it out? Click here for a a free trial.