What you can automate in a Tienda Nube store with conversational AI: products, orders, coupons, shipping, and how per-store learning works.
ES · Read this post in the other language PT · Read this post in the other languageRunning a Tienda Nube store means repeating the same administrative work every week: loading products, adjusting prices, checking orders, creating coupons, updating shipping settings. Each task is small, but together they consume the hours you would rather spend on the parts of the business only you can do. This guide explains what conversational AI can realistically take over, how it learns your store, and how to start without risking your catalog.
The idea behind conversational e-commerce automation is simple: instead of clicking through admin panels, you describe what you want in plain language, and the system executes it against your store using structured, pre-built procedures. You say what should happen; the AI figures out which operations to run.
With Cangrejito, a Tienda Nube store can delegate work across products, orders, customers, coupons, shipping, categories, and marketing, backed by more than 55 pre-built operational procedures. In practice this covers most of the recurring administrative work of store operations: creating and updating products with variants, answering questions about order status, generating discount coupons with specific rules, adjusting shipping configuration, and keeping categories organized. The AI does not improvise raw actions against your store; it composes tested procedures, which is what makes delegation safe enough for daily use.
Product work is where conversational automation saves the most time. You can ask for a new product with several variants in one message, request price adjustments across a set of items, or ask which products are missing descriptions or images. Because the AI reads your existing catalog first, new entries follow the same naming and structure you already use.
Order operations are mostly about answering questions fast: which orders are pending, which ones are paid but not shipped, what a specific customer bought. Instead of filtering panels, you ask. The AI returns the answer and can take follow-up actions where a procedure exists for them.
Coupon creation is a classic example of a task that is simple but tedious in an admin panel. Conversationally, you describe the rule, for example a percentage discount with a minimum purchase and an expiry date, and the coupon is created with exactly those parameters. This becomes especially useful for recovery and campaign flows.
Shipping settings and category structure change less often, but when they change, they tend to change in bulk. Reorganizing categories or reviewing shipping options conversationally avoids the repetitive clicking that makes these tasks error-prone.
Cangrejito builds an operations manual specific to your store. From the first connection, the AI studies your catalog structure, your naming conventions, your shipping rules, and the way you run promotions, and it records what it learns as reusable knowledge inside your isolated tenant. Every task it completes refines that manual. The result is that the tenth task is more precise than the first: the AI already knows how you abbreviate product names, which categories matter, and what a normal discount looks like in your store. That knowledge never leaves your tenant and is never shared with other stores.
This matters because generic automation fails on specifics. A rule that works for one store's catalog breaks another's. Per-store learning is how conversational artificial intelligence adapts to your business instead of forcing your business to adapt to it.
Delegation without visibility is a bad trade, so Cangrejito surfaces everything through kanban task boards and metrics dashboards. Work the AI takes on appears as cards you can review, and dashboards show what was executed and what changed. You decide how much autonomy to grant: start by reviewing everything, and loosen the reins as the operations manual matures and results stay consistent. You can see the full list of covered areas on the features page.
The plan is flat: USD 20 per month per installation, including 10,000 emails per month, 2,000 chat conversations per month, and 5 generations. Overage is USD 0.001 per additional email and USD 0.02 per additional conversation. There are no usage tiers and no per-seat pricing, so the cost of automating your store operations is known before you start.
Starting takes minutes, and the sensible path is gradual:
Automation is not a switch you flip; it is a working relationship you build. The stores that get the most out of conversational AI are the ones that start small, verify results, and then delegate more each week. A Tienda Nube store that follows that path typically ends up handling its routine operations in conversations that take seconds, with the admin panel reserved for the exceptions.
