*By Jorge Moskovitz
News reports have been highlighting that Artificial Intelligence development companies have begun directly offering customized consulting services for AI implementation. But why have these companies, which are among the most advanced in development, decided to invest in services and implementation, areas traditionally reserved only for large technology consultancies?
This question opens the door to strategic reflection: Are they changing their business model? Will they begin to compete directly with consulting firms that support large corporations in adopting technologies? Or are they simply recognizing where the true competitive advantage has always been? The answer is simple: they understand that AI isn't the product. It's the execution!
The advantage has never been access to baseline models. It has always been the ability to integrate, adapt, and operationalize these models in a realistic, continuous, and aligned manner with the business strategy. This is precisely what generates value in the corporate world.
We live in an era where APIs, Large Language Models (Large Language Models – LLMs) and algorithms have become commodities. Almost all businesses have access to the same technologies, developed by so-called big tech companies. But if the models are widely available, why are the results still so uneven? This is because the difference lies not in the technology itself, but in how it is applied, customized, maintained, and integrated into workflows and business decisions. In other words, it lies in the execution, and this is transformative.
For a long time, the prevailing narrative suggested that simply adopting a generative AI API was enough to position oneself at the forefront of innovation. But the reality is different. According to a McKinsey report, nearly 1,001% of companies invest in AI, but only 11% consider themselves mature in terms of implementation. The difficulty of integrating AI into workflows, with practical use and supported by executive leadership, is among the biggest obstacles.
Data-native companies have always understood that the challenge wasn't simply "buying AI." True value arises when AI is operationalized with reliable data, in a scalable, ethical manner, and integrated with business objectives. This is why some AI solutions, even ready-to-use platforms, come with implementation services. Without quality data, well-defined pipelines, and clear governance, there is no value delivery or transformation.
Implementing AI without planning is like putting a Formula 1 engine into a popular vehicle without adapting the chassis, gearbox, and brakes. And then expecting it to run smoothly on the streets of our cities without a professional driver. This engine change would bring power, but it's useless without the proper infrastructure. For AI to work, it needs a track (strategy), drivers (skilled professionals), fuel (well-managed data), and clear signage (governance). Only then can it cross the finish line well-positioned in an increasingly competitive and fast-moving market.
We are witnessing an inevitable repositioning in the market. "AI as a product" is out, and "AI-driven execution as a service" is in. This requires a technical, cultural, and strategic shift from organizations, as it's not just about innovating, but about doing so consistently. Developing an API is now a simple and inexpensive task, but keeping it alive within a complex ecosystem of corporate data and ensuring it is up-to-date, secure, and governed is a major challenge.
AI only generates value when built on strong foundations of data quality, efficient orchestration, and cost control. Therefore, companies should invest in partnerships with established AI-based solution providers who are focusing on Agentic AI initiatives and the integration of different AI solutions to meet this goal. They should align with those who are enabling autonomous agents capable of interacting with structured and unstructured data, APIs, and corporate processes, always with guaranteed human oversight and auditability. And, above all, they should provide highly consultative support from initial architecture to production support.
Artificial Intelligence that transforms is the one implemented, integrated, and managed correctly. The question every company should be asking itself today isn't which AI model to use. In fact, there are several other more important ones right now: "Are we buying an AI solution as a final product or as the beginning of a transformation journey?"; "Who will be by our side when obstacles arise?"; "Who will point out flaws in our data and help us fix them?"; "How will we keep our APIs alive, secure, and adapted to the constant changes, rules, and regulations of the market?"; and "How will we adapt our operational flows to the AI logic—and not the other way around?"
Responding to and addressing these questions in practice is what separates projects that generate real returns from those that get lost in pretty presentations and failed deliveries. After all, AI It's not the product of a company seeking to use it. It's just another means to bring, elevate, and accelerate the real product to market. Those who can truly implement and execute it will come out ahead.
And you, what are you doing about it?
*Jorge Moskovitz, Enterprise Account Executive at Qlik
Notice: The opinion presented in this article is the responsibility of its author and not of ABES - Brazilian Association of Software Companies
EN 
PT 
