A customizable agentic AI toolkit for e-commerce

Build and deploy AI agents grounded in product data and workflows. Open-source, extensible, and ready to run in minutes.

GithubDocs

Core Features

Deploy to cloud or on-premises

Run Enthusiast locally with Docker Compose or scale it in production on Kubernetes. Maintain full control over infrastructure and data.

Model-agnostic architecture

Use OpenAI, Azure, Gemini, Mistral, or Ollama. Swap models through configuration without altering application logic.

E-commerce integrations included

Native connectors for Shopify, Shopware, Medusa.JS, and Solidus. Synchronize product catalogs, content, and documentation without additional integration work.

Pre-built agents for common workflows

Ready-to-use agents for product search, catalog enrichment, content generation, and support Q&A. Extend or adapt them to match your processes.

Customizable by design

Built with Python (Django), React, and LangChain. Extend with Python plugins, custom tools, or bespoke agents without forking the core platform.

LangSmith tracing integration

Enable execution traces by setting environment variables. Send agent runs to LangSmith out‑of‑the‑box to review operations and improve prompts and workflows.

Enthusiast provides everything you need to begin building e-commerce AI workflows.

Engineered for teams that require transparency and control over their AI systems.

Open-source

Modular

Extensible

AI-based Product Discovery

Enable natural-language product search. Users describe requirements in plain terms, and the agent retrieves verifiable results from your indexed catalog.

User Manual Search

Provide precise answers from technical documentation and manuals. Retrieval-augmented generation ensures responses are accurate and grounded.

OCR to Order

Convert unstructured inputs — such as scanned invoices or PDFs — into structured orders. Automate validation against your catalog before passing data downstream.

Product Catalog Enrichment

Generate product descriptions, attributes, or translations directly from your data. Validation agents confirm factual accuracy prior to publication.

Technology Ecosystem

Technology Stack

Integrations

Supported models

Build with confidence

Deploy your first agent in under five minutes.

Get Started Schedule a Demo

What Technical Leaders Ask Us

How does Enthusiast integrate with existing e-commerce stacks and legacy systems?

Enthusiast ships with native connectors for Medusa, Shopify, Shopware, and Solidus, but it’s built to extend. You can connect to ERP, CRM, or inventory systems via APIs, middleware, or custom plugins in Python. This allows you to run AI workflows without refactoring your core commerce stack, and ensures compatibility with platforms like SAP, Salesforce, or custom databases.

Can I control which AI models are used, and swap them without rewriting logic?

Yes. Enthusiast is model-agnostic by design. You can configure it to run on OpenAI, Azure, Gemini, Mistral, or Ollama. Switching models is as simple as updating configuration—no changes to business logic or agent code. This ensures future-proofing against vendor lock-in.

What deployment options are available for production workloads?

You can run Enthusiast locally with Docker Compose for quick prototyping, or deploy to Kubernetes for high-availability production environments. Both cloud and on-premises setups are supported, giving you full control over infrastructure, scaling, and compliance requirements.

How does Enthusiast ensure transparency and control over generated outputs?

All workflows are grounded in your product catalog, documentation, or structured data. Retrieval-augmented generation ensures factual accuracy, and validation agents can be added to enforce business rules. This reduces hallucinations and keeps outputs consistent with your brand and compliance standards.

Is Enthusiast extensible beyond the pre-built agents?

Absolutely. While the toolkit ships with ready-to-use agents for product search, catalog enrichment, support Q&A, and order processing, you can extend it with custom Python plugins, LangChain tools, or fully bespoke agents.

How secure and compliant is Enthusiast for enterprise environments?

Since you can deploy on your own infrastructure (cloud or on-premises), all sensitive data stays under your control. Enthusiast supports GDPR and PCI DSS compliance requirements and allows you to integrate with existing authentication, logging, and monitoring stacks.

Will Enthusiast always remain open source?

Yes. Enthusiast is released under an open-source license and is designed to remain community-driven. The core toolkit will always be available to developers at no cost. Upside maintains the project publicly on GitHub, with contributions and governance open to the community, ensuring long-term availability and transparency.

Does Upside provide consulting or enterprise support for Enthusiast?

Yes. While the toolkit is open source, Upside offers consulting and custom implementation services for advanced use cases. This includes system audits, integration with legacy platforms, large-scale deployments, and the development of bespoke agents or connectors. For technical leaders who require guarantees on delivery, scaling, or compliance, Upside’s consulting team ensures successful adoption and measurable business results.