- Lohika solved critical architectural problems which improved code maintainability, increased code coverage, added performance metrics to all application critical components.
- Developed performance tests to verify SLAs, and created an integrated development environment powered by Docker and Docker Compose.
- Lohika ensured reliability and monitoring by migration of core components to Apache Airflow. Critical parts of the system were moved to Kubernetes cluster in Google Cloud.
- Lohika improved reporting system test coverage and corrected significant performance problems.
- Next step is to move essential reporting to a new data platform to improve data availability and accuracy.
- Lohika omitted existing authentication solution in favor of Keycloak. This replacement improved scalability, availability, and security.
- Lohika introduced DevOps standards and tools including Jenkins, Zabbix, NewRelic and also Jenkins Groovy SDK, which allowed engineers to build CI/CD pipelines without involving DevOps in future.
Making Advertising More Valuable to the World
Blending data science, objective media and captivating experiences
Challenge
- Evolve its core data spine project, but it was hard to develop, maintain, and scale.
- Improve its core ETL and data warehousing system by increasing application performance and scalability.
- Tune a highly customizable reporting system by improving data availability and correctness.
- Client needed to develop a custom authentication and authorization solution and improve functionality related to scalability, availability, and security.
- Introduce modern software development practices and functionality, including a DevOps culture, mindset, and modern tools.
Overview
Our client is part of the world’s largest marketing & digital insights group that blends data science, objective media and captivating experiences to build valuable connections between brands and consumers.
Solution
Results
- Introduced a brand new data platform, defined architecture that leverages industry standard tools (Talend, Hive, Presto, Looker, etc.) and hosted in GCP.
- Drastically improved overall app performance, modernized build infrastructure and created reproducible development environment.
- Automated deployment of core components and applications. Introduced Kubernetes as a primary path for all new deployments.
