Artificial intelligence at the heart of your catalog
AI Copilot, multi-phase enrichment, semantic search, DeepL translation, compliance extraction — 19+ providers and services configurable with your own API keys.
Query your catalog
in natural language
No more SQL queries or technical training. Ask your question in plain English — Copilot translates it into a database query and answers instantly.
- →Questions about your catalog, your imports, your completeness scores
- →Complex queries (multi-filters, aggregations, comparisons)
- →Natural-language answers with the underlying figures
- →Accessible to your whole team, no training required
Example questions
How many products have no description in my Electronics category?
Catalog audit
Which products have a completeness score below 60%?
Data quality
Show me the 20 most expensive products without a main image.
Missing media
Which suppliers delivered imports with errors this month?
Import tracking
List in-stock products that are non-compliant with REACH.
Compliance
What is my completeness progress by category?
Dashboard
collect → synthesize pipeline with RAG
Not just a "GPT that writes". A structured pipeline that first queries your supplier documents (local Qdrant RAG — zero cost), falls back to web search if needed, and synthesizes everything into a publish-ready sheet.
Collect — RAG + Web
Local Qdrant · Perplexity / SearXNG
The pipeline first queries your indexed document base (RAG on local Qdrant — supplier technical sheets, manuals, certificates). If RAG coverage is sufficient (≥ 3 relevant chunks), no web call is made — zero AI cost. Otherwise, automatic fallback to a configurable web search (Perplexity or self-hosted SearXNG).
Synthesize — LLM
OpenAI / Anthropic / Mistral / Ollama…
An LLM of your choice consolidates documentary data (primary source) and web data (secondary source) into an enriched sheet: SEO description, extracted technical attributes, dimensions, weight, image alt tags, competitor prices and official videos. Documentary sources always take priority over web.
19+ providers and services — your keys, your costs
Bring your own API keys. Your AI costs stay under control and transparent — no separately billed credits, no black box.
| Provider | Type | Primary use |
|---|---|---|
| OpenAI | LLM + Vision | Enrichment, Copilot, Compliance, Vision |
| Claude (Anthropic) | LLM + Vision | Enrichment, Copilot (fallback) |
| Google Gemini | LLM + Vision | Enrichment, Vision |
| Mistral AI | LLM | Enrichment, Copilot |
| Grok (xAI) | LLM | Enrichment |
| Groq | Fast LLM inference | Ultra-fast enrichment |
| DeepSeek | LLM | Enrichment |
| Together AI | LLM gateway | Multi-model, open source |
| Ollama (local) | Self-hosted LLM | 100% private, no data sent out |
| Ollama Cloud | Hosted LLM | Managed self-hosting |
| llama.cpp | Local C++ LLM | Ultra-fast local inference |
| OpenRouter | Multi-LLM gateway | Unified access to 100+ models |
| DeepL | Translation | Product sheet translation (BYOK) |
| Web search | Web search | Enrichment phase A (configurable) |
| Web search (self-hosted) | Web search | Enrichment phase A (self-hosted) |
| AI web search | AI web search | Enrichment phase A (configurable) |
| Firecrawl | AI web scraping | Structured web data collection |
| Jina AI | Embeddings + Search | Semantic search, reranking |
| Shopping API | Google Shopping | Enrichment phase B (prices, sheets) — configurable |
| Icecat | Product data | Specialized enrichment |
| Amazon SP-API | E-commerce | Listings, ASINs, Amazon catalog |
| Amazon Advertising API | Ads / Shopping | Amazon advertising campaigns |
Ollama enables a 100% local deployment — your data never leaves your infrastructure.
All AI capabilities
Automatic translation (DeepL)
DeepL BYOK integrated — translate your product sheets into multiple languages with your own API key. Batch translations across your entire catalog, no hidden surcharge.
Semantic search
Qdrant embeddings to find similar products by meaning. "Waterproof sports shoes" also finds "running trail waterproof".
Vision & AI tagging
Automatic analysis of your product images. Alt tag generation, semantic tagging, visual anomaly detection.
Compliance extraction
Analyzes your product documents (PDFs, datasheets) and automatically extracts CE, REACH, RoHS, energy and certification data.
Provider Assignments
Configure which AI provider for which context: enrichment, copilot, translation, vision, compliance — cost/quality optimization.
Custom prompts
Prompt library by product category. Define your tone, your required attributes, your SEO rules — reuse across your entire catalog.
Document RAG — zero cost
Index your supplier technical sheets, manuals and certificates in local Qdrant. The enrichment pipeline uses them as the primary source — no web LLM call while documentary coverage is sufficient.
Smart Router & Cost Estimator
11 AI contexts assignable to different providers (enrichment, copilot, translation, vision, compliance, embeddings…). Response caching, cost estimation before execution, automatic fallback on error or rate limit.
MCP — external AI agents (dev / power-users)
8 MCP (Model Context Protocol) servers expose 42 JSON-RPC tools to connect your catalog to Claude Desktop, IDEs or custom AI agents. Dev capability, not a UI feature.
Ready to enrich your catalog with AI?
Start for free — no credit card required.