WooCommerce AI Sales Agent: 11 features most chatbots don’t have (and why they matter)
Ask most AI shopping assistants a simple question – “how many colors do you carry this jacket in?” – and you’ll often get a confident, wrong answer. That’s not a rare glitch. It’s structural: semantic search retrieves individual products, and individual products can’t be added up into a count. A WooCommerce AI sales agent built to answer that class of question correctly, rather than around it, is a different product from a chatbot that just talks well.
Corelex has eleven capabilities like this already built and largely unmarketed – verified against the codebase, not the pitch deck. The two worth leading with: an aggregate-question engine that solves a problem most competitors don’t even know they have, and a search-box substitution that goes head-to-head with dedicated standalone search plugins at a fraction of the cost. Here’s the full list.
Key takeaways
Your catalogue changes by the minute – does your AI know within seconds, or within a week?
Corelex updates its Vertex AI Search index within seconds of any WooCommerce product change – a save, a stock update, or a deletion triggers an automatic sync immediately. Unpublishing a product is treated the same as removing it, so it stops appearing in recommendations right away, not at the next sync cycle.
How fresh that data is at the moment of a query also depends on plan. Starter and Growth stores work from a short cache – 60 seconds by default – which comfortably covers normal catalogue activity. Professional plan switches to a fully real-time mode instead, bypassing the cache and reading price and stock straight from WooCommerce on every request. That difference matters most around time-sensitive changes – a flash sale, a limited restock, a same-day price correction – where even a minute of staleness can mean quoting the wrong price or recommending something that’s already sold out.
The gap this closes is a familiar one: a customer asks about a product that was unpublished an hour ago, and a chatbot running on a weekly or nightly sync recommends it anyway. That’s not a hallucination in the usual sense — it’s stale ground truth presented with full confidence.
Not convinced of the added value of an AI Sales Agent?
The question every RAG-based shopping assistant gets wrong
Ask a typical AI shopping assistant an aggregate question – how many colors do you carry, what’s your price range, how many products are on sale right now – and it will often answer confidently and incorrectly. Semantic search retrieves individual matching products; it has no mechanism to sum, count, or range across them. Corelex closes that gap with a facet engine that computes these answers directly from catalogue data, not from whatever handful of products retrieval happened to surface.
|
Question type |
Typical RAG-based assistant |
Corelex |
|---|---|---|
|
“How many colors does this come in?” |
Approximates from the products it retrieved, often undercounting |
Computed from catalogue facets – exact |
|
“What’s your price range for hiking boots?” |
Guesses from a sample, not the full range |
Computed across the full matching set |
|
“Show me the cheapest option” |
Ranks by relevance, not price – approximate |
Direct price sort against live WooCommerce data – exact |
That last row is its own feature: a “cheapest” or “most expensive” query routes straight to a direct WooCommerce price-sort call rather than an approximation built from retrieved products. The answer is exact because it comes from the same sort WooCommerce itself would run – not a best guess dressed up as a fact.
This is the single easiest way to see the difference between a chatbot and a sales agent in a live demo: ask any competing tool an aggregate question, then ask Corelex the same one.
Security checks that live in the code, not the prompt
Corelex’s order-lookup logic refuses to run unless the customer’s email has been verified. That check is enforced in application code, not delivered as an instruction the model is expected to follow. It’s a meaningful distinction: a system-prompt instruction can be argued with, reframed, or talked around by a sufficiently determined user. Code that simply doesn’t run the lookup without verification has no such surface to attack.
The same code-first approach extends across the platform’s security posture:
No SMB-tier WooCommerce AI competitor currently documents this level of architectural detail publicly. It’s the kind of thing that matters most to the buyer who asks the hardest questions – an agency vetting a tool for a client’s flagship store, or an IT lead sizing up AI vendors for the first time.
A drop-in replacement for your WooCommerce search box
Corelex can intercept the native WooCommerce or WordPress search widget and route each query into the same hybrid retrieval used in chat conversations. The result: a store’s existing search bar starts returning meaning-aware results – matching intent and synonyms, not just exact keywords – without the customer ever opening the chat widget.
For merchants currently paying for a dedicated search plugin, this is a direct substitution, not an add-on feature. Standalone WooCommerce search tools with AI-enabled tiers commonly price in the $150–350/month range. Corelex includes conversational search replacement inside its existing plans, alongside everything else the sales and service agents do — one subscription instead of two separate tools solving overlapping problems.
One agent, two jobs: sales and support, with or without WooCommerce
Corelex runs in three modes: Sales Agent, Service Agent, or both together. The Service Agent works entirely on its own – answering from WordPress pages, posts, and uploaded PDFs – with no WooCommerce connection required at all.
That widens who Corelex is actually for. The “WooCommerce plugin” framing is accurate but incomplete: any WordPress site with documentation, policies, or product information spread across pages and PDFs can run Corelex as a knowledge-base assistant, whether or not it sells anything through WooCommerce. A services business, a membership site, or a content publisher can use the same underlying agent a webshop uses for product questions.
Ready for the EU AI Act disclosure rule that’s already in effect
Since August 2, 2026, Article 50 of the EU AI Act requires that anyone interacting with an AI system – a chatbot included – be clearly told they’re talking to AI, not a person, at the first point of contact. This isn’t a future compliance item on a roadmap. It’s a current legal obligation, with fines reaching up to €15 million or 3% of global turnover for non-compliance.
Corelex already ships an AI-disclosure badge on every chat widget, visible from the first message. Merchants running Corelex don’t need a separate project to meet this requirement – it’s already there.
Frequently asked questions
What is a WooCommerce AI sales agent, and how is it different from a chatbot?
A chatbot answers conversationally, a sales agent takes the same conversation further – it can compute exact answers (counts, ranges, price sorts), verify identity before showing sensitive data, and take actions like adding a product to cart. The distinction matters most on questions a chatbot can only approximate.
How fast does Corelex know when a product’s price or stock changes?
Within seconds. WooCommerce updates trigger an immediate sync to the Corelex search index. On Professional and Enterprise plans, price and stock are also read live from WooCommerce on every request instead of a short cache – useful for flash sales or fast-moving stock.
How does Corelex answer questions like “how many colours do you carry”?
Corelex uses a facet engine that answers aggregate questions – counts, ranges, totals – by computing directly from catalogue data rather than approximating from a handful of retrieved products. This is the type of question most AI shopping assistants answer incorrectly.
Does Corelex verify a customer’s identity before showing order details?
Yes. Order lookups require a verified email address, and that check is enforced in the application code itself – not left as a instruction the AI model is simply expected to follow.
Can Corelex replace my WooCommerce search bar?
Yes. Corelex can intercept the native WooCommerce or WordPress search widget and route queries into the same conversational retrieval used in chat, so search results understand intent and synonyms rather than requiring exact keyword matches.
Does Corelex work as a support agent even without WooCommerce?
Yes. The Service Agent mode answers from WordPress pages, posts, and PDFs on its own, with no WooCommerce connection required – useful for any WordPress site, not only webshops.
Is Corelex ready for the EU AI Act’s chatbot disclosure rule?
Yes. Corelex ships an AI-disclosure badge on every chat widget by default, meeting the Article 50 transparency obligation that became enforceable on August 2, 2026.
Where is my store’s data stored, and how is it isolated from other merchants?
Merchants choose their data region – EU, US, or Global – at setup. Each merchant’s data lives in its own per-tenant Vertex AI Search datastore, and backend access is locked to the API Gateway’s service account only, so no other tenant’s traffic can reach it.
The through-line
Catalogue freshness, aggregate answers, exact price sorts, and code-enforced identity checks aren’t four separate features – they’re the same feature applied four times: Corelex is built to be right, not just fluent. That’s a harder thing for a competitor to fake in a sales demo than a longer feature list, because it shows up the moment someone asks a question a typical assistant can only guess at.
See the full feature set on the Corelex product page, or check the numbers for your own store on the Corelex ROI page.
