- All the infrastructure layer are done internally. We don't use any of the cloud public offering like AWS or Azure. Not sure how this will scale in the future.
- Lack some structure and process around scrum management and more generally around engineering and management.