Floom is an AI tool designed to orchestrate and execute generative AI pipelines. It is conceptualized as a 'Kubernetes (K8s) for AI', providing an environment for developers and DevOps to focus on their core tasks. Functions of Floom encompass a variety of tasks, including data ingestion, cost control, and caching. With Floom, users can establish AI pipelines that apply models, set constraints, and include validation checks and response formats. Its compatibility with different models and APIs, such as OpenAI-GPT 3.5 and DALL-E, offers users flexibility in their AI deployments. Additionally, Floom provides multiple packages to tailor pipeline functionalities, such as privacy filters for personal details, bad-words filter for profanity, cost management for controlling expenditure, and the cache package for efficient storage management. These packages promote the secure, efficient, and responsible use of AI. Floom also offers get-started and quick-start documentation that provides easy-to-follow guides for users. It is open-source, encouraging developers to contribute and adapt its functionalities to their specific needs.