The global market of MLOps (Machine Learning Operations), solutions that help data scientists to simplify and optimize machine learning deployment processes, will have an average annual growth of almost 45% by 2030. The projection was made by the research company Valuates Reports, which estimates a jump in the valuation of the segment of US$ 186.4 million, achieved in 2023, for US$ 3.6 bi. One of the main reasons for the warming of this market may be in the reduction of the term for the development of predictive models. The evaluation of the company'specialized value of Datar is Data Reliefs in the use of the company'sdecision as a service”.
To develop similar systems with traditional methods, organizations take an average of two to three weeks, depending on the complexity of the industry.
“On the other hand, by using MLOps the data scientist can automate the entire creation process. First he does all the training part of the model through an automatic machine learning that tests algorithms to see which one works best. At this time, the scientist can also, if he wants, upload a code that he already has his own and save all the documents and all the codes, thus ensuring the protection of the documentation of all the databases. The success of MLOps is due to the fact that he eliminates all these steps with the creator of the model being responsible himself and having in hand everything he needs to go from the beginning to the end of the” project, he says.
In 2024, Datarisk launched to the market a MLOps solution focused on serving leading companies in activities such as credit granting, fraud risk, propensity to change work, productivity in the agro, among others. Only during the first half of this year, the tool was used to perform a volume of more than 10 million queries and, among the benefits obtained by users of this technology, one of the greatest highlights was precisely the reduction of time. With the startup MLOps, the average term of three weeks fell to a matter of hours.
Carlos Relvas also explains that, after this first training is built, a second step enters within the Datarisk MLOps platform itself, which is the part in which the scientist can automatically, himself, create an API for the model to be used in external environments. The third step, according to him, is the management of the solution. At this stage, the objective is to ensure that this model that has been developed, trained and is being used continues to have a good performance over time.“A tool can monitor both the use of its applications and the operation of APIs to ensure not only that everything is operating as scheduled, but also allow the measurement of the quality of the model to be changed, the solution is changing itself.
The market receptivity and the prospects that Datarisk has made allow the company to project growth of more than five times the volume of use of this solution by the end of 2025.
The Co-founder and CEO of Datarisk, Jhonata Emerick, explains that by becoming a pioneer in offering solutions in the MLOps concept in Brazil, the startup is putting into practice the strategy of maturing and perfecting its main business theses. “We understand in greater depth the market needs and are now prepared to offer solutions capable of transforming in an absolutely relevant way the reality of data science in the country”, he says.
According to Emerick, in the specific case of predictive model development, MLOps solutions emerge as a response to time-consuming internal processes designed for an era when companies did not have to manage a data area with the agility that is currently required.
“It is generally adopted the IT queuing systems in which the data science area finishes making a model and passes to the engineering area to create an API. This, in turn, will take a significant time to do its part, when then the project will pass to the credit engine team, for example, so that it finally implements this API, which will lead to other deadlines. The result is that when the model is implemented, the situation is already different. Therefore the MLOps solution becomes so effective in terms of optimization,” concludes.
