Rivet - Page-aware AI shopping agent for ecommerce product pages
byā¢
Rivet Chat is an AI shopping assistant for Shopify and custom stores. Page-aware product cards, live catalog answers, and add-to-cart attribution. Embed in about 2 minutes.
Replies
Best
Maker
š
Rivet Chat is an AI shopping assistant that lives on ecommerce product pages. It detects which product a shopper is viewing, answers stock and pricing questions by querying the store's live catalog (not a static FAQ), shows a product card with an Add to cart button inside the chat, and tracks which add-to-cart events came from the conversation.
It installs on Shopify via App Embed, or on custom React/Next.js storefronts via the npm package smartreply-chat. No model training ā merchants upload a CSV or sync products, and the assistant answers from current data.
Report
Maker
Hi Product Hunt! š
I'm the maker of Rivet Chat. Here's the quick version of why we built it:
Most ecommerce chatbots give wrong inventory and pricing answers because they read from a static FAQ document uploaded months ago. The moment you upload it, the data starts going stale.
Rivet queries the store's live product database on every message. So when a shopper asks "Is this in stock?", the answer comes from the actual catalog ā not a training file.
A few things that make it different:
ā¢ Page-aware ā detects which product the shopper is viewing and shows a product card for that item
ā¢ Add to Cart inside the chat, tracked via ATC attribution
ā¢ Works on Shopify (App Embed, no code) + custom React/Next.js (npm: smartreply-chat)
I'm here all day ā ask me anything about the architecture, the live-DB-vs-RAG tradeoff, or how ATC attribution works. Happy to do technical deep dives. š ļø
Replies
Hi Product Hunt! š
I'm the maker of Rivet Chat. Here's the quick version of why we built it:
Most ecommerce chatbots give wrong inventory and pricing answers because they read from a static FAQ document uploaded months ago. The moment you upload it, the data starts going stale.
Rivet queries the store's live product database on every message. So when a shopper asks "Is this in stock?", the answer comes from the actual catalog ā not a training file.
A few things that make it different:
ā¢ Page-aware ā detects which product the shopper is viewing and shows a product card for that item
ā¢ Add to Cart inside the chat, tracked via ATC attribution
ā¢ Works on Shopify (App Embed, no code) + custom React/Next.js (npm: smartreply-chat)
I'm here all day ā ask me anything about the architecture, the live-DB-vs-RAG tradeoff, or how ATC attribution works. Happy to do technical deep dives. š ļø