Serviette is a collection of ready-to-deploy cloud templates for building retrieval-augmented generation and AI enterprise search applications with live data synchronization.
The tool addresses the challenge of building production-grade RAG systems that maintain accuracy by staying synchronized with current data. It provides pre-built application templates that automatically sync with multiple data sources including file systems, Google Drive, Sharepoint, S3, Kafka, PostgreSQL, and real-time APIs. The templates include built-in data indexing supporting vector search, hybrid search, and full-text search, all performed in-memory with caching. Users can deploy templates as-is or modify individual pipeline steps, such as swapping a vector index for a hybrid index, through simple configuration changes.
Developers should choose this tool if they need to deploy RAG or enterprise search applications quickly without managing separate infrastructure dependencies. The templates are designed to scale to millions of documents and work across Linux and macOS environments. The approach suits teams wanting to avoid building indexing and synchronization logic from scratch, particularly those whose data sources span multiple platforms. The tool is Docker-friendly and supports deployment to cloud providers including GCP, AWS, and Azure, as well as on-premises environments.
The project maintains active community engagement through Discord and social channels. Development focuses on providing templates optimized for different use cases, ranging from simplicity to high accuracy. The codebase is structured around Jupyter Notebooks as the primary format, making templates accessible and modifiable for users familiar with notebook-based development workflows.
