Blog, Pitang.

Society is facing a series of disruptive changes that have radically transformed the way we perform tasks. The fast pace of these transformations drives the emergence of new technologies, expanding the frontiers of our knowledge. In the same way that electricity boosted factory work beyond what was thought possible during the industrial revolution, the internet enhances the capabilities of a single computer. 

The transition from paper-based bureaucratic processes to digital operations has driven digital transformation across various industry sectors. With the advances in this transformation, online applications are no longer an exclusive need of large enterprises, becoming indispensable for any digital product or company. 

Public cloud providers

Maintaining an online application in a self-hosted environment is quite complex and costly. This challenge drove the growth of third-party services that offer online application hosting, as well as the emergence of public cloud providers. It is important to emphasize that a public cloud provider does not necessarily offer free services. The term "public" simply means that anyone can use their cloud services to host an application. 

Currently, cloud providers offer a broad and constantly expanding range of services, covering computing, data storage, networking, and artificial intelligence services. The main public cloud providers are Amazon Web Services (AWS), Google Cloud Platform (GCP), and Microsoft Azure. 

Cloud and DevOps

As software engineering advanced, the widespread use of cloud computing contributed to the popularization of agile methods in software development. This trend gave rise to the DevOps culture, also known as development and operations.

DevOps is a software development approach in which deploying code to the cloud is no longer the developer's sole responsibility, but is automated through a set of services.  

DevOps uses the principles of continuous integration (CI) and continuous deployment (CD). Continuous integration enables automated code testing and improvements, catching errors before deployment, while continuous deployment refers to shipping the code to the cloud after testing and integration.

Currently, all major cloud providers have DevOps pipelines that assist in implementation, compatible with other widely available open-source tools. The availability of these pipelines transforms the way we build and deploy code, providing significant gains in terms of productivity and the reliability of code delivered to production. 

Artificial Intelligence and Machine Learning

Currently, applications utilizing artificial intelligence (AI) and machine learning (ML) are becoming increasingly popular across various digital products. Companies are realizing the value of these approaches and beginning to implement them in their production environments. However, developing, testing, deploying, and improving these digital products is more complex than developing a standard web page or mobile app. 

ML systems are composed of different parts, such as the code that implements the algorithm, the model generated by that code, and the data used to train and evaluate the model. These parts are always subject to change and must be carefully managed. 

Furthermore, it is important to highlight that running a machine learning system on a personal computer differs from running it in a production environment. The behavior of these systems is harder to predict, test, improve, and explain when in production. 

MLOps: machine learning operations

With the increased demand for machine learning systems to be available online and in production, needs similar to those that led to the development of DevOps have emerged. In recent years, the term MLOps, which stands for machine learning operations, has been used to describe the extension of DevOps to the field of ML systems. 

In short, MLOps is an approach that aims to facilitate the development, testing, deployment, and improvement of machine learning systems, ensuring they operate effectively and reliably in production environments. 
 

Pitang innovates through Pitang Labs

Pitang, as a company committed to innovation and the development of advanced technological solutions, has been investing heavily in research and experimentation through Pitang Labs.

This in-house laboratory is dedicated to exploring disruptive technologies, including the use of machine learning in the cloud and in digital products. We believe in the transformative power of these approaches and are thrilled to share our knowledge and experience with you.

Don't miss our article series that will cover in detail how to use machine learning in the cloud and in products, keeping up with the industry's latest trends and best practices. Stay tuned and join us on this exciting journey toward technological innovation!