Artificial intelligence and software testing: Oculow

With a solid foundation in Artificial Intelligence and software testing, Oculow fills us with pride. It is an IT company that was created as a spinoff of Abstracta to apply Artificial Intelligence (AI) in software testing, through our Abstracta Revolution program.

Abstracta Revolution is a program that encourages people who work in our company and its close community to create technology ventures, startups and spinoffs. To undertake in technology and turn their ideas into sustainable, profitable and scalable businesses.

Oculow is undoubtedly a great example of all this, of what it means to be a technology entrepreneur and of the importance of Artificial Intelligence in the future of testing. Its creator is Diego Ferrand, and he works as a programmer at Abstracta.

“Oculow is a tool that allows automated tests to be extended to perform visual validation. It detects image overlapping errors and also has functionalities that already exist in other image comparison tools. While it has the image comparison part, this is somewhat complementary to demonstrate how it can be used in synergy with similar tools”, introduced Diego. 

The main objective of the spinoff Oculow is to “improve the AI part”. So that visual validations can be performed with better accuracy and find other types of errors, such as element alignment and design anti-patterns, among others. 

The road to this point was full of challenges, including a variety of validations and exchanges. “Abstracta supported me with mentoring during the development process and during the product commercialization stage”, he recounted. 

“Abstracta introduced me to contacts through which we were able to secure funding for the project. It made it easier for me to reach a variety of companies, to approach them with the product and obtain input. It gave me access to the service of automated testers to use the tool and contribute to its improvement”, he continued.

After applying to Abstracta Revolution to become a technology entrepreneur and create a spin-off, he got feedback from the board of directors. And, gradually, everything began to take shape. 

He started having weekly meetings with Abstracta’s Quality Engineer and CEO Matias Reina, as a mentor for the project. He took the idea to several companies that work with Abstracta to get more feedback and validate the market for the tool. And, finally, he managed to strengthen ties with the business incubator Gepian in Salto, Uruguay. 

Through Gepian, the possibility arose to present the project to the National Agency for Research and Innovation (ANII) to obtain seed capital and speed up the development of the product. This process lasted 1 year and it was worth it. He obtained $20,000 in funding from the ANII to support his spin-off! 

He quickly began the development of the Minimum Viable Product (MVP), i.e. the simplest version of the product, as a tool to test his business idea. During this development stage, the team expanded to 3 developers and the brand definition was achieved. Later, when the MVP was finished and the product was ready to be used, the sales stage began. 

– How was the idea of Oculow born?

I was always interested in machine learning. The visual validation tools on the market today work by comparing images. This means that you need to define a base image and then compare future versions of the application with that base image. In other words, if you want to test in different resolutions, you need to review and define the base image for each screen you want to validate. This makes it very difficult to scale the tests. In addition, any minor differences, such as language or pixel shift differences, will cause the tests to fail. Seeing all these difficulties that exist in the visual validation part, I thought that maybe some machine learning could be applied to facilitate and support these tools.

– What motivated you to sign up for Abstracta Revolution?

I have been working at Abstracta as a programmer since 2016. One day, I was talking with the CEO Matías Reina about how Artificial Intelligence can be applied in the testing part. A truly combination of Artificial Intelligence and software testing. I told him that it would be good with a tool capable of identifying visual errors automatically, and that the results of this would be notified to the developers without the need to manually review each resolution. Mati thought it was a good idea and told me about Abstracta Revolution. There I started the whole process of entrepreneurship and in 2019 I made the first steps.

– What is the current status of Oculow?

Today the tool can be used to develop visual tests, being able to test AI technology and see how it can be applied for base image definition for more traditional systems. Currently, I am working on improving and expanding the detections that are performed. One of the biggest changes I am implementing is to use AI with programmatic rules. This will help in the accuracy of the detections as it can rule out cases where, for example, an image that has intentionally overlapping elements. It will also help speed up the implementation of future detections. 

– Is Abstracta still present in all these processes?

Of course! Abstracta continues to support me with a large network of contacts and infrastructure. It supports me with automators who use the tool as part of learning, giving feedback on improvements that can be applied to the tool.

– Is it common for companies to drive this type of actions?

I don’t know of any other company that has any initiative similar to this, not by far! Abstracta Revolution is an incredible resource for anyone interested in technology entrepreneurship. 

– How did starting this project change you professionally?

The whole experience helped me to learn and understand all the stages that a project takes when starting a tech startup. I managed to grow as a developer thanks to the technical challenge of developing a product from scratch. I also managed to strengthen many of my soft skills, which as developers costs us a lot. In addition, I gained new perspectives: I began to see and understand tools and companies from a different angle. 

– What would you say to someone who plans to start a technology business? 

Entrepreneurship does not come with a guide of the steps to follow, it is a unique path for each person. If we don’t have someone to guide us and support us in difficult moments, it is easy to lose motivation in the project. One of the most important things when starting a business is networking, it is easier to get a meeting with someone when you are introduced to them than to do it from scratch. My experience at Abstracta is amazing. I started on June 1, 2016 as an automator and shortly after that I started my developer stage within the company. It was really my first work experience and the opportunities that have opened up for me I could never have imagined. Entrepreneurship in technology is a path of tremendous learning, and I hope to continue to grow much more in it. 

Artificial intelligence and software testing

We live in a globalized and completely interconnected world. As such, the code of our future is marked by the IT world and Artificial Intelligence and software testing tools. Artificial intelligence and software testing are crucial in this future. Combined with the professional empowerment of people, erasing boundaries and expanding all kinds of horizons. 

Technology entrepreneurship and spinoffs will undoubtedly be a major cornerstone of this future.

Would you like to apply to Abstracta Revolution? Contact us at [email protected]

Follow us on Linkedin, Facebook, Twitter, and Instagram to be part of our community!