First Round was LLM based Refactoring Assessment and had to do 5 tasks with help of AI but highly encouraged to do documentation and write tests yourself, I felt the time was too less to do 5 tasks in an unknown repo and just using AI to do all 5 problems wasn't the task I guess