Karrot Tech Stack
Community super app with secondhand marketplace, groups, and local services
Karrot operates a multi-service community platform across Korea, Japan, the US, Canada, and the UK, with over 30 million registered users and 20+ million MAU. The tech stack reveals a data-and-ML-forward architecture: BigQuery, Apache Airflow, dbt, TensorFlow Extended, and Kubeflow power recommendation and ad-targeting models, while mobile-native tooling (RxSwift, Alamofire, fastlane) supports iOS-first product design. Hiring velocity is decelerating despite open roles; the org currently balances data engineering (3), product engineering (3), and single-IC coverage across design, finance, HR, marketing, ops, and sales—suggesting either stable operations or selective scaling paused post-Series D.
Tech Stack 54 technologies
What Karrot Is Building
◆Challenges
- Lack of differentiation clarity
- Low product utilization
- Slow performance diffusion
- Improving user engagement
- Enhancing ad targeting accuracy
- Preventing incident recurrence
- Building incident response system
- Detecting security incidents
- Scalable architecture for super app
- Maintaining large-scale services
▲Active Projects
- Custom sales strategy planning
- Network security architecture
- Agency discovery expansion
- Personalized recommendation system improvement
- Advertising targeting model improvement
- Building security platform
- Building incident response system
- Consulting desk operation
- 중고거래실
- 모바일실
