The talk describes what challenges HERE faces in processing Data Subject Requests, e.g. data deletion requests, and how we automated this process to scale it to process up to thousands of requests per week. The talk introduces an experimental approach that uses LLMs to automated extraction of relevant information from free-text emails.