The Experience Assistant is an AI assistant built directly into Icecat PIM, designed to answer questions about your product catalogue using live catalogue data and extensive platform documentation. Whether you’re looking for a specific product, troubleshooting an import or checking whether your catalogue is ready for export, the assistant provides accurate, contextual answers without leaving your workflow.
Managing product information means working across thousands of products, attributes, imports, exports, pricing rules and sales channels. Finding the right answer often means searching documentation, checking dashboards or asking a colleague.
Powered by Claude, Anthropic’s AI model, the Experience Assistant operates entirely within Icecat PIM. It uses the same catalogue data and improved documentation your team already relies on, making it a natural extension of your existing workflow rather than another external AI tool.
We’re rolling it out in stages, and this page will remain the central place to discover new capabilities as they become available.
Current capabilities
Ask questions the same way you would ask a colleague.
The Experience Assistant analyses your live catalogue together with Icecat PIM’s documentation to provide contextual answers tailored to your environment. Because it connects to nearly every part of your PIM through the platform API – including products, dashboards, imports, exports and more – it answers based on the current state of your catalogue rather than a static help article.
Today, the assistant can retrieve information, analyse your catalogue and explain how Icecat PIM works. It also understands and responds in multiple languages, allowing international teams to work in the language they are most comfortable with while sharing the same catalogue and workflows.
The next stage will introduce more capabilities, allowing the assistant to perform actions on your behalf after approval. You’ll find more information about that further down this page.
See the Experience Assistant in action
The best way to understand the Experience Assistant is to see it at work.
Icecat PIM generated the genuine conversations below. They showcase just a small selection of what you can ask the assistant – from getting started with the platform and understanding advanced functionality to analysing your own catalogue.
In practice, the Experience Assistant answers questions about your own data. You can ask about products, brands, pricing, stock levels, imports, exports, dashboards, enrichment, integrations, documentation and much more. If the information exists within Icecat PIM, the assistant helps you find and understand it.
We did not rewrite or simplify these responses for communication purposes. They are authentic outputs from the Experience Assistant. The examples use representative demonstration data. We include any brands, distributors, product counts or catalogue information shown solely to demonstrate what the assistant can do. They do not represent a customer environment.
🤖 Experience Assistant: Welcome! Here’s a friendly overview of what Icecat PIM is and where to start.
What is Icecat PIM?
Icecat PIM is a Product Information Management system built for e-commerce and distribution. It’s the single place where your team can:
- Create & enrich products — build your catalogue, add descriptions, images, videos, attributes and categories. Icecat PIM can also automatically pull in rich content from the Icecat database, with AI-assisted enrichment and translation.
- Manage pricing — set purchase and sell prices per warehouse and condition, configure VAT rates, currencies and pricing rules.
- Import data — bring in supplier and distributor product feeds, daily price files and Icecat content on schedules or on demand.
- Export & distribute — send your catalogue out as files (CSV, XML, etc.) or through direct connectors to platforms like Magento, Shopify, WooCommerce, Shopware, Amazon, Google Merchant Center and more.
- Manage orders — a built-in order layer lets you track sales and purchase orders alongside your catalogue.
Who typically uses it?
| Role | What they do |
|---|---|
| Catalogue / content managers | Build the catalogue, enrich products and manage translations |
| Pricing managers / buyers | Maintain prices, VAT, currencies and pricing rules |
| Integration / operations | Configure imports/exports and monitor jobs |
Where to begin
Here are the recommended first steps:
- Explore the Dashboard — After logging in, your dashboard gives you live KPI widgets: product counts, enrichment status, channel activity, recent imports/exports and more.
- Check your Products — Browse your existing catalogue, or start creating products. You can filter by brand, category, warehouse and more.
- Set up Categories & Brands — Organise your catalogue structure so products are easy to find and export cleanly.
- Configure Imports — If you have supplier feeds, set up an import to bring products in automatically.
- Set up Exports / Channels — Connect to your webshop or marketplace and start distributing product data.
- Use the built-in Help module — Go to Sidebar → Help inside the interface for guided walkthroughs, video tutorials and release notes.
Feel free to ask me anything specific—whether it’s looking up products, checking your dashboard statistics, managing brands or configuring imports and exports. I’m here to help!
Sources
- User Guides › Getting Started › Overview
- User Guides › Sources › Imports
- User Guides › Exports › Template Exports
- User Guides › Access › Login
- User Guides › Help › Help
🤖 Experience Assistant: Great question! Here’s a thorough breakdown of both filter types and how to use them well.
The Import Pipeline
Every import flows through these stages in order:
File → Pre-filter → Mapping / Naming Conversions → Post-filter → Write to catalogueUnderstanding where each filter sits is the key to using them correctly.
Pre-filters — The First Gate
What they do:
Pre-filters look at each incoming row before any mapping, enrichment or naming conversions happen, and decide whether to let the row through or drop it immediately.
Critical rule:
They operate on raw, supplier-side values — the warehouse code, brand name, category path and product type exactly as they appear in the feed. If you write a canonical PIM name here, nothing will match.
Best used for:
- Cutting a large supplier feed down to only what you actually sell (specific warehouses, brands, categories and product types)
- Blocking junk rows — test SKUs, archived brands and irrelevant categories — before they waste processing time
- Speeding up imports by skipping work on rows you’d discard later anyway
How rules work:
- Multiple rules combine with OR — to allow brands A, B and C, create three separate rules.
- Fields within a single rule combine with AND — e.g. brand = “Acme” AND warehouse = “EU01”.
- An empty field means “all” (no constraint on that dimension).
- Wildcards/LIKE syntax are not supported — use the exact supplier value.
Behaviour options:
Each pre-filter set is either Include (only matching rows pass) or Exclude (matching rows are dropped).
⚠️ If you switch a set to Include without adding any rules first, every row will be dropped.
Post-filters — The Final Gate
What they do:
Post-filters run after mapping and naming conversions are complete, so each row already carries fully resolved, canonical PIM data (real brand IDs, real category IDs, real prices, stock levels, etc.).
Best used for:
- Decisions that depend on canonical/mapped values, e.g. “drop products without a recognised brand”
- Filtering by stock level, price range and Icecat content availability
- Keeping only products in specific PIM categories (including sub-categories)
Post-filter rules are much richer than pre-filter rules — you can filter on:
- Category (with sub-category inclusion and wildcard matching)
- Brand, Product type, Warehouse, Channel, Condition and Definition
- Purchase price range, Stock level range and Restock interval
- Whether the product has Icecat content
- Whether the EAN is complete
Multiple rules use OR; fields inside one rule use AND — the same logic as pre-filters.
Useful tip:
Use the Show selected products button on the post-filter list to preview exactly which products would be in or out of scope before running an import.
Side-by-Side Summary
| Pre-filter | Post-filter | |
|---|---|---|
| When it runs | Before mapping & conversions | After mapping & conversions |
| Values it sees | Raw supplier values | Canonical PIM values |
| Use case | Drop irrelevant/junk rows early | Filter on resolved PIM data |
| Rule fields | Warehouse, Brand, Category, Product type | Category, Brand, Stock, Price, Channel, Content and more |
| Wildcard support | ❌ No — exact match only | ✅ Yes — wildcard mode available |
| Preview tool | Import Insights (after the run) | “Show selected products” (live preview) |
Practical Tips
- Use pre-filters first, post-filters second. Pre-filters are cheaper — they skip processing entirely for dropped rows.
- Check the logs. If rows go missing, open Export/Import Insights — both filter types record exactly which rule dropped each row and why.
- Reuse pre-filter sets. The same set can be attached to multiple imports.
- Auto-create pre-filter rules from your Naming Conversions — the Create prefilters bulk action on the brand/category conversion list generates matching rules automatically.
Sources
- User Guides › Sources › Import Pre Filters
- User Guides › Sources › Post Filters
- User Guides › Sources › Import Pre Filters
- User Guides › Sources › Import Pre Filters
- User Guides › Sources › Import Pre Filters
- User Guides › Sources › Imports
- User Guides › Sources › Imports
- User Guides › Sources › Post Filters
🤖 Experience Assistant: Here’s your catalogue readiness overview across 19,007 products:
✅ Product Status — Excellent
| Status | Count | % |
|---|---|---|
| Online | 18,993 | 99.93% |
| Expected | 9 | 0.05% |
| Offline | 2 | 0.01% |
| Customized range | 2 | 0.01% |
| End of Life | 1 | 0.01% |
Nearly your entire catalogue is Online — great shape for an export.
⚠️ Category Coverage — Needs Attention
- 17,860 products (94.0%) have at least one category assigned.
- 1,147 products (6.0%) have no category — these may display poorly or be filtered out in your webshop depending on its setup.
⚠️ Enrichment — Mostly Good
- 18,205 products (95.8%) are Icecat-enriched ✅
- 802 products (4.2%) are not yet enriched — they may lack descriptions, specifications or images.
- No manually enriched products on record.
📦 Last Export — Recent
The most recent product export was a JSON run completed on 15 July 2026 at 14:41.
The last CSV product export was on 7 May 2026.
🔑 Summary & Recommendations
Your catalogue is largely export-ready, but there are two gaps to address before the next webshop export:
- 1,147 uncategorized products — assign categories so these products are correctly navigable in your webshop.
- 802 un-enriched products — review these for missing content (descriptions, images and specifications).
Would you like me to dig deeper into either of these issues—for example, listing the uncategorized products or filtering by un-enriched ones?
Coming soon
Today, the Experience Assistant reads, analyses and explains.
The next stage will allow it to act on your catalogue by creating products, adjusting data, and carrying more complex tasks.
Every action will remain behind an approval step, ensuring that no changes are made without human confirmation.
As new capabilities become available, we will update this page with additional examples and use cases.
How the Experience Assistant differs from a general-purpose AI chatbot
A general-purpose chatbot answers using public information and general knowledge. The Experience Assistant combines Icecat PIM’s documentation with your own live catalogue – so it answers questions a generic chatbot simply cannot.
Grounded in your live catalogue: It analyses your products, imports, exports, dashboards, pricing, channels and catalogue data through the Icecat PIM API – not generic advice.
Connects information across your PIM: It pulls from multiple parts of Icecat PIM – your catalogue, dashboards and documentation – to explain what happened and why, rather than surfacing a single help article.
Built on extended Icecat PIM documentation: Its explanations track how the platform actually works, not assumptions.
Secure by design: It runs in an isolated environment and draws only on data you’ve authorised it to access. Your catalogue data is never used to train any AI model.
Honest when information is unavailable: When something can’t be determined, it says so and points you to the relevant documentation or data source. For product information, transparency beats a confident guess.
The Experience Assistant builds on Icecat PIM’s existing AI-powered capabilities for product enrichment and follows the same principles outlined in our Do’s and Don’ts for AI in PIM: use AI to improve product information and automate repetitive work, while keeping humans in control of important decisions.
Why an AI assistant for Product Information Management matters
The Experience Assistant turns “where does this information live?” into a single question. Instead of learning where everything sits, your team just asks for it.
That shift – from experimenting with AI to embedding it directly in daily work – is where the research points. OpenAI found employees using AI assistants save 40–60 minutes a day; the London School of Economics put the average at 7.5 hours per week. McKinsey tracks that same move from experimentation to embedding AI in day-to-day workflows – exactly the role the Experience Assistant fills inside Icecat PIM.
As a result, Product Information Management teams spend less time searching for information and more time improving product data.
Get started
The Experience Assistant is rolling out across Icecat PIM and will continue to gain new capabilities over time.
Whether you’re exploring Product Information Management for the first time or looking to make your existing catalogue easier to manage, the Experience Assistant helps you find answers faster without leaving your workflow.
Ready to see it in action?
- Start exploring yourself in our Free Trial of Icecat PIM, or
- Book a demo with our team, or
- Learn more about Icecat PIM
Elisabeth is part of the MarCom team at Icecat Commerce, where she combines technical product knowledge with real-world client situations. Having worked directly with clients on their product data challenges, she writes from the perspective of what actually matters in day-to-day e-commerce operations.
