Croct uses Google Cloud to deliver a unique customization experience in real time
About Croct
Croct is a SaaS company engaged in the customization of applications in real time. Using a SDK, the company provides a solution for developers to create natively customized apps, as well as for marketing and product professionals to offer unique experiences to users based on their behaviors, interests and browsing history. In short, it offers a method to transform data into actions in real time.
Tell us your challenge. We're here to help.
Contact usThanks to Google Cloud, the startup built the infrastructure for its platform, as well as tools that provided the necessary scalability and performance for the solution.
Results
- Reaches latency levels below 5 ms to set up a full user and session assessment context.
- Scales services at a rate of thousands of events per second.
- Reclaims ~30% of savings thanks to the migration of databases to Cloud SQL.
- Eases application management with Google Kubernetes Engine.
Deployed data streaming infrastructure in 4 months.
Customization of sites and apps based on different user profiles is a strategy increasingly used by companies willing to optimize lead capture or increase its online conversion rates. Understanding the behaviors and characteristics of target audience segments, it is possible to establish differentiated and more efficient ways to communicate with each of them through customized elements in images, texts, colors, buttons and more.
However, for many organizations, this strategy demands a huge investment in development and even the adoption of several tools to perform different actions. Moreover, a large portion of these solutions only enables customization outside the app, which narrows the options and makes usage more difficult. This means that, in certain situations, it is necessary to have the entire site developed in the first place to apply customizations later with the aid of an external tool - and to adapt to the options it offers.
Croct was created in 2016 with the aim of making this process more practical and scalable within companies. The idea was to provide a single solution that not only enabled and simplified the native creation of customizations (from the onset of app development) but also offered resources such as A/B testing, data analytics, recommendations, and more.
Over four years, the startup studied this problem in depth and built a set of tools to ease the work of development, product and marketing professionals. The team even created its own coding language, CQL (Contextual Query Language), using English as a reference to make the insertion of customizations easier.
All this required robust technology and a flexible environment. Croct chose Google Cloud to support the company’s architecture.
“We decided to start operating with Google Cloud to leverage scale from day one, using a cloud infrastructure so as to avoid the need of managing physical servers. We chose Google Cloud because of all the managed services it offers and our trust in the experience of Google in its applications.”
—Marcos Passos, founder, CroctChoice of tailor-made solutions
At the beginning, working with a small team of engineers, Croct assessed which Google Cloud services the company would use to select those which really satisfied its needs. “We always had our environment in mind, we didn’t want a ready-to-use infrastructure where we could insert our application and that’s it. We always focus on generating quality code to test and scale and, likewise, we needed a well-designed infrastructure to do everything more efficiently,” Luis Ferraz, DevOps engineer at Croct explains.
Ferraz, having worked in a Google Cloud partner before, led the deployment of the solutions, which took four months to complete. The main tools used were: Google Kubernetes Engine (GKE), to run applications; Cloud SQL, to manage databases; BigQuery, to store data, and Cloud Storage, to store assets, backups and historical data.
Thanks to the integration between Croct’s architecture and Google Cloud’s infrastructure, the startup was ready for product launch, which finally took place in 2020. The benefits of using Google Cloud solutions quickly surfaced.
A scalable, flexible and easy-to-manage environment
The ease of application management provided by GKE and the little need of intervention were two strengths highlighted by the team. Flexibility came as a total surprise by furnishing cost optimization opportunities within the tools. By migrating GKE databases, previously managed directly by Croct, to Cloud SQL, managed by Google Cloud, the startup achieved 30% cost and time savings, which helped the company focus more on strategies and processes to add value for end users.
“We detect areas with a potential for optimization and generate savings, both in relation to Kubernetes, enabling us to automatically scale down and reduce the number of machines during low-traffic hours, and also to managed applications.”
—Luiz Ferraz, DevOps engineer, CroctThese tools also contributed to the performance of the solution offered by Croct. Cloud SQL enabled latency levels below 5 ms to set up a full user and session assessment context. Also, GKE enabled service scaling at a rate of thousands of events per second.
In any data-driven company, with infrastructure as one of its pillars, maintaining a good level of proximity with its cloud provider is critical.
“Having a single point of contact for communications is a huge differential of Google Cloud. The proximity with our account manager speeds up everything. We have learned to value this aspect of support, mainly because other providers do not offer this.”
—Marcos Passos, founder, CroctFerraz also highlights the vast documentation and community engagement level, as the startup can solve queries and study tools and integrations in depth. As an example, the DevOps engineer mentioned Google Cloud articles that aided the startup in the use of tools through Google Magic Modules to define the entire infrastructure. This enabled to simplify the performance of tests to solve issues, optimize the use of resources, and select the best tools, as it is possible to replicate the in-use infrastructure in different testing environments separately.
A step forward with machine learning
Assessing the use of Google Cloud tools did not only occur at deployment. The team checks for new possibilities to adopt new solutions or optimize the existing ones on an ongoing basis. Some managed services options are already in the startup’s radar when higher-scalability needs arise.
Moreover, Croct is working in a project using machine learning solutions. This project aims at improving user experience and helping customers in the generation of revenue using reinforcement learning. “Google Cloud is at the forefront of machine learning research and development efforts, and has a huge infrastructure to make everything easier. That’s why we will surely use all the features of AI Platform a lot in the next few months,” Passos says.
Tell us your challenge. We're here to help.
Contact us
About Croct
Croct is a SaaS company engaged in the customization of applications in real time. Using a SDK, the company provides a solution for developers to create natively customized apps, as well as for marketing and product professionals to offer unique experiences to users based on their behaviors, interests and browsing history. In short, it offers a method to transform data into actions in real time.