*Per rodney disgust

The efficiency of Artificial Intelligence depends on its access to data systems that are correctly integrated. This is what a recent survey conducted by Salesforce points out –, which revealed that only 28% of applications, on average, are integrated with each other, with 95% of IT leaders reporting integration issues that are hindering the adoption of Artificial Intelligence.

The study reinforces that successful AI strategies that drive return on investment (ROI) and business value depend on the integration and unification of data that can reap the benefits of improving operational efficiency, productivity and customer experiences. employees and customers.

IT leaders expect a 69% increase in the average number of Large Language Models (LLMs) they will use over the next three years, with 80% of organizations reporting they already use multiple predictive and generative AI models today.

Other findings from this research reinforce this need for application and data systems integration projects: 81% of IT leaders state that data silos are hindering digital transformation efforts; 62% report that their data systems are not configured to fully leverage AI; and that project requests from the IT team increased by around 39% in the last year.

Breaking integration breaking down data silos

For a long time, data silos have been considered one of the main significant barriers to digital transformation and the consolidation of business value. In this survey, 81% of respondents revealed that silos are hindering their digital transformation efforts and, as a result, there is a marked need to improve integration to unify all business data – structured with unstructured –, aiming to leverage and deploy reliable, business-relevant AI.

However, organizations are not equipped to harmonize their data systems to take advantage of AI technologies, a situation highlighted by 62% of respondents. Another two-thirds of IT leaders (72%) also consider current infrastructure to be overly interdependent, with 62% of them reporting that their organization is not equipped to harmonize its data systems to take advantage of AI technologies. The result: Data insights remain untapped, with 75% of organizations struggling to integrate data insights into user experiences.

In turn, customers' digital experiences are not fully prepared: only 26% of organizations believe they provide a completely connected user experience across all channels.

Where does the data come from, where is it and where is it going?

According to the International Data Corporation (IDC), the majority of data generated worldwide (60%) comes from companies, originating from sales, purchases, payments transactions, from interactions with customers via websites, applications and social media. Added to this is data collected from IoT (Internet of Things) sensors, equipment monitoring, as well as documents, reports, emails, presentations, contracts, etc.

All this volume of data is transported via different applications and systems, such as ERP, CRM, SCM, BI, BPM, HR, CMS, e-Commerce, internet banking and many more.

The isolated data in each of these systems, without any communication between them, results in the inability of Artificial Intelligence to act directly on them. For this reason, the success of AI strategies depends on their ability to access the structure of structurally integrated systems.

Commenting on the survey data, Param Kahlon, Executive Vice President & General Manager, stated: “This data shows that IT leaders are increasingly aware that their ability to operationalize AI — and generate business and customer value — will depend on their ability to integrate it with their existing systems to create new growth opportunities.”

(*) rodney disgusted, CEO of Magic Software Brasil.

Notice: The opinion presented in this article is the responsibility of its author and not of ABES - Brazilian Association of Software Companies

quick access