Simple AI application embeddings database
Build LLM applications without the complexity of setting up vector databases
Perfect for serverless apps, prototyping and more. Dead simple API gets you started in seconds
RagTag provides a simple, quick database to store document embeddings for LLM applications accessible over a REST API
Store and retrieve documents using simple API calls. No need to learn about the intricacies of vector databases, embeddings, chunking and distance search. RagTag APIs handle the complexity of converting your documents to vectors and returning the most relevant content to queries.
Retrieve relevant data from your data collections for RAG, recommendation, detection and other applications. RagTag is built on the open-source Chroma vector database, which prioritizes speed and high performance.
Make all your data available to your AI. RagTag scales to support your application needs whether it is one document or 10,000.
RagTag supports all sizes of mission-critical applications. Support SLAs available.
LLM platforms (ChatGPT, Anthropic, Gemini, etc) are trained on publicly available data and may have no context about your specific use cases. Retrieval Augmentation Generation (RAG) is a technique to provide the LLM information relevant to the query as needed. This provides the LLM the relevant context to use in its response and reduces the risk of hallucination.
RagTag stores all your data in collections. These collections include not only the content but also the metadata (eg section names, urls, titles, etc) that can provide a richer knowledgebase for the LLM. These documents are automatically converted to embeddings and stored ready for vector operation.
RagTag collections can be queried to provide relevant documents and their associated metadata. These results can then be provided to the LLM as context to the query.
RagTag is in invite-only mode right now. Please provide the details of your project and use case and we will get back to you as quickly as possible. In the meantime, get familiar with the API documentation