Enthusiast

Enthusiast is an open source agentic AI toolkit for ecommerce, published by the software company Upside. It gives engineering teams a framework for building and deploying agents that are grounded in a live product catalog, store content and technical documentation, rather than a general model with no view of the merchant's data. Teams can run it locally with Docker Compose or scale it in production on Kubernetes, keeping infrastructure and data under their own control. Several agents ship ready to use. Product search takes an unstructured request and narrows the catalog by asking targeted follow up questions until it reaches a short list. User manual search answers from documentation using retrieval augmented generation. Catalog enrichment generates descriptions, attributes and translations, with validation agents confirming accuracy before publication. Catalog web import parses product pages from supplied URLs and writes the results into the catalog in a single batch. Invoice scanning extracts stock keeping units and quantities from supplier PDFs and scans, updating stock levels and flagging unmatched items for review. Native connectors cover Shopify, Shopware, Medusa, Solidus, WooCommerce, WordPress and Sanity. The architecture is model agnostic, so OpenAI, Azure OpenAI, Gemini, Mistral, Anthropic or a local Ollama instance can be swapped through configuration. It is built with Python and Django, React and LangChain, extends through Python plugins and custom tools without forking the core, and sends execution traces to LangSmith when environment variables are set. No pricing is published. The toolkit is distributed as open source on GitHub with its own documentation site, and the parent company's pricing page is not publicly reachable, so any commercial support or implementation work is arranged directly with the vendor. The documentation lists release 1.7.0 as current.