Kognia: Video Annotation Tool & SDK Frontend Architecture
Kognia Sports Intelligence is a Barcelona-based startup redefining tactical football analysis through AI-driven, real-time and post-match insights. Unlike traditional metrics such as heat maps or distance traveled, Kognia focuses on identifying high-level tactical patterns, guided by a team of experts led by Joan Vilà, former FC Barcelona Director of Methodology. Their vision goes beyond clubs: they aim to provide 360º insights to scouts, medical teams, broadcasters and other sports professionals. To extend their platform, they needed a new manual video annotation tool to complement their automated analysis engine.
Kognia faced a complex challenge: developing an SDK-style frontend architecture within an environment not originally designed for such modularity. Although the initial plan was to build the new tool as an isolated project, system dependencies —including authentication, user management and internal dashboard flows— made that approach unfeasible. The challenge was to create fully reusable, typed React components that could operate inside a large monorepo with multiple teams working in parallel, ensuring stability, compatibility and seamless integration across different areas of the platform without disrupting existing workflows.
Adidas’s mobile development journey
The collaboration began with the development of SDK components to deliver essential features for Adidas’s applications. This included highly customized UI elements, modules for network communications, and customer chat functionality integrated into the SDKs.
In parallel, Jenkins build and deploy jobs were set up to automate delivery and improve efficiency in the development lifecycle. These efforts strengthened the continuous integration pipeline and reduced release times.
The project also introduced Docker and Kubernetes to reinforce the infrastructure, providing greater flexibility in managing environments and a more scalable foundation for future developments.
Finally, Grafana was integrated to monitor backend systems and logs, combined with issue management practices to detect problems early and ensure a more stable service.
Technology
Technologies used: Android & iOS native development (Java, Swift, Objective-C), Docker, Kubernetes, Grafana
Results
- Development of SDKs supporting all Adidas applications on iOS and Android
- Delivery of key features for the newest Adidas mobile app
- Improved QA processes across devices and applications
- Standardization of components and processes across projects
- Enhanced infrastructure with CI/CD, Docker, Kubernetes, and monitoring tools
We are ready for new challenges
