Are data and AI projects slow (also) in your company? So slow that you don’t even want to start one.
Over the past year, we’ve noticed a big shift in our customers’ data needs – or, more specifically, the business needs to use data. The availability and development of AI technologies has opened up new opportunities for data exploitation, but has also created challenges for businesses in delivering projects.
Why is development so slow?
While the number of ideas has increased and new technologies are more available, the practical implementation of ideas is often slow. The most common reason is simple: data is not readily available and it takes a long time to start using new data.
Imagine you go to the grocery store to buy basic groceries. You don’t want to spend hours talking to the shopkeeper about the store layout, the selection or the healthiness of the products. The assumption is that the products are fresh, safe and the label quickly tells you what they contain.
The same logic should apply to data. Typically, our customers have identified dozens if not hundreds of ideas where data and AI could be used. The number of ideas is huge, but traditional implementation methods do not allow for rapid execution of ideas.
The same jars work for different recipes
The use of AI requires that data is available and of high quality. We’ve solved the problem by working with our customers to develop data mesh type of data management architectures where data is commoditized like canned tuna on the store shelf. The same data can be used in business development data projects as well as regulatory projects
Five steps to accelerate data projects
1. Create a portfolio of data needs, as it is more cost-effective to develop together than separately.
2. identify the data you need and flag the ones that are most in demand.
3. Use the lessons of product management to productize the data.
4. Develop data management in a use-case oriented way, i.e. do not create management processes before there is a real need for them.
5. Seek support from someone who has already done data productization before.
