Shoponix AI Chat for WooCommerce
This plugin adds an advanced AI chatbot to your WooCommerce store that answers customer questions and recommends products. It uses Retrieval-Augmented Generation (RAG) to ensure the AI only answers based on your actual store inventory. Powered by the Mistral AI API, it securely stores product embeddings directly in your WordPress database, requiring no third-party vector databases. Features: – Conversation Memory: The AI remembers context, allowing customers to ask follow-up questions seamlessly. – Real-time Streaming (SSE): AI responses stream instantly word-by-word just like ChatGPT. – Rich Product Cards: When the AI recommends a product, a visual WooCommerce product card (with image, price, and link) is rendered directly inside the chat. – Interactive Product Comparison Card: Customers can request a side-by-side spec and price comparison table of multiple items inside the chat widget. – Visual Image Search: Users can upload images to find visually similar products in your store. – Conversational Checkout: Users can add to cart, choose shipping, and securely checkout completely within the chat window. – Order Modification & Cancellation: Allows users to cancel or modify eligible orders directly via chat. – Context-Aware Browsing History: The AI tracks recently viewed products to provide highly relevant contextual responses (e.g., “Does this come in blue?”). – Proactive Sales Recovery: Features an Exit-Intent popup and proactive Abandoned Cart recovery prompts offering tailored discounts. – Text-to-Speech (TTS): The chatbot can read its responses aloud with a sleek UI toggle. – Sentiment-Based Live Escalation Desk: Monitors client mood to dynamically surface a premium human support ticket form whenever a user gets frustrated. – First-Screen Language Selection: Includes floating visual toggles (supporting English, Hindi, Spanish, French) that dynamically translates the UI and forces the AI response to localise naturally. – Dynamic Variation Selector: Replaced static selectors with adaptive attribute selectors (supporting Color, Size, Fit, etc.) dynamically syncing variation matrices. – Custom Quantity input: Added standard capsules along with an inline custom numeric quantity option. – Background Indexing: Uses WooCommerce’s Action Scheduler to index products in the background without causing timeouts, regardless of your store size. – Incremental Updates: Product vectors automatically update in real-time when you edit, create, or delete products. – Inventory Awareness: The AI knows if products are In Stock or Out of Stock. – Floating chat widget with customizable colors and shortcode support [shoponix_chatbot]. – Rate limit handling with automatic retries for the Mistral API. External Services This plugin relies on external third-party APIs to generate product vector embeddings and power the conversational AI chat interface. Depending on the provider configured in the plugin settings, data (such as product titles and descriptions for indexing, or user/cart context and chat transcripts during a chat session) is sent to one of the following services: Mistral AI Service URL: https://mistral.ai Terms of Service: https://mistral.ai/terms-of-service/ Privacy Policy: https://mistral.ai/privacy-policy/ Data Sent: Product titles and descriptions for generating embeddings; customer chat history to generate chatbot responses. OpenAI Service URL: https://openai.com Terms of Service: https://openai.com/policies/terms-of-use/ Privacy Policy: https://openai.com/policies/privacy-policy/ Data Sent: Product titles and descriptions for generating embeddings; customer chat history to generate chatbot responses. Google Gemini Service URL: https://ai.google.dev Terms of Service: https://developers.google.com/terms Privacy Policy: https://policies.google.com/privacy Data Sent: Product titles and descriptions for generating embeddings; customer chat history to generate chatbot responses.
Top keywords
- ai15×2.75%
- chat11×2.02%
- product11×2.02%
- https9×1.65%
- chatbot6×1.10%
- mistral6×1.10%
- products6×1.10%
- responses6×1.10%
- service6×1.10%
- com5×0.92%
- embeddings5×0.92%
- mistral ai5×0.92%
AIPCF – AI Provider for Cloudflare
This plugin connects WordPress to Cloudflare Workers AI, making its hosted large language models available to any WordPress feature or plugin that uses the WordPress AI Client (introduced in WordPress 7.0). Supported models The plugin fetches the full list of available text generation models directly from your Cloudflare account and shows them in the settings dropdown, so you always see what is actually available to you. The list is cached for 12 hours and refreshed whenever you save your credentials. A built-in fallback list is used when your credentials are not yet configured or the API is unreachable. See Cloudflare’s model catalogue for the current full list. Why Cloudflare Workers AI? Unlike hosted AI services that charge per seat or per API call at a premium rate, Cloudflare Workers AI runs inference at the edge on a usage-based model. The free tier covers 10,000 neurons per day (Cloudflare’s billing unit — roughly 200–500 average generation requests) — enough for a content-heavy site to generate post summaries, meta descriptions, and moderation checks without any cost. Paid usage is billed per neuron, making it significantly cheaper at scale compared to per-token pricing from closed providers. Your prompts and completions also never leave Cloudflare’s network to reach a third-party AI company’s servers, which matters for sites with privacy or data-residency requirements. What you need A Cloudflare account with Workers AI enabled. Your Account ID — visible on the Workers AI dashboard sidebar. A Cloudflare API token with the “Workers AI – Edit” permission. Usage Once the plugin is configured, any WordPress AI Client-compatible feature or plugin will automatically use Cloudflare Workers AI as its provider. Developers can also call it directly: use WordPress\AiClient\AiClient; $result = AiClient::prompt( 'Summarise this post in two sentences.' ) ->usingProvider( 'cloudflare-workers-ai' ) ->usingModel( '@cf/meta/llama-4-scout-17b-16e-instruct' ) ->generateTextResult(); echo $result->toText(); Switch models without changing any other code — swap llama-4-scout-17b-16e-instruct for qwen2.5-coder-32b-instruct and you are running a coding assistant instead. This plugin does not add a chatbot or front-end widget by itself. It registers the provider so that other AI Client-powered features, plugins, and themes can use it. External Services This plugin connects to the Cloudflare Workers AI REST API to run AI inference. The service is used to: Run text generation requests against the model you select. Verify your Account ID and API token when you click “Test connection”. What data is sent and when When you click “Test connection” in the plugin settings, your Account ID and API token are sent to https://api.cloudflare.com/client/v4/accounts/{account_id}/ai/models/search to verify the credentials. No prompt data is sent during this check. When a compatible AI Client feature generates text, the plugin sends your API token, the selected model ID, and the prompt messages to https://api.cloudflare.com/client/v4/accounts/{account_id}/ai/run/{model_id}. Requests are only made when AI generation is explicitly triggered — nothing is sent automatically in the background. Service provider Cloudflare Workers AI is provided by Cloudflare, Inc. Cloudflare Workers AI documentation: https://developers.cloudflare.com/workers-ai/ Cloudflare Terms of Service: https://www.cloudflare.com/terms/ Cloudflare Privacy Policy: https://www.cloudflare.com/privacypolicy/ Roadmap Current version supports Workers AI edge inference. AI Gateway support for caching, rate-limiting, and analytics coming soon.