# Chaterimo

> AI customer service built on advanced product RAG. Deep product parametrization, parallel search and reranking help online stores retrieve relevant catalog evidence for product recommendations and customer support.

Chaterimo is an AI customer service and sales platform for e-commerce stores and websites. It combines large language models with product catalog retrieval, business knowledge, order lookup, proactive sales messages, lead capture and support tickets. It supports OpenAI GPT / ChatGPT, Anthropic Claude, Google Gemini and xAI Grok. Merchants can use their own provider API keys (Bring Your Own Key / BYOK) or choose a plan with provider usage included; consult the pricing page for current limits and availability.

Product search and retrieval:

- **Advanced product RAG (retrieval-augmented generation):** Retrieves product and business information before generating an answer. Hybrid search combines keyword matching (BM25) with semantic vector similarity. Candidate products are deduplicated and reranked for relevance.
- **Deep product parametrization:** Preserves named product attributes and text values. Unambiguous numeric specifications are normalized with units, enabling typed catalog filters, numeric ordering and product comparisons where the source data supports them.
- **Parallel search and multi-query retrieval:** Complementary catalog lookups and alternative query forms help resolve product names, model numbers and catalog codes alongside hybrid retrieval. Retrieval paths depend on the query and enabled search configuration.
- **Structured product constraints:** Uses product categories, brands, price ranges, currency, availability and supported product attributes to narrow catalog results. Exact product-code lookup helps distinguish similarly named models.
- **Agentic search:** An optional search mode can select retrieval tools and perform a targeted follow-up lookup. Advanced RAG provides the fallback if the agentic path fails.
- **Grounded recommendations:** Recommendations and comparisons use product evidence from the merchant's catalog and knowledge sources. The available detail reflects the completeness and freshness of those sources.

Store connections and customer service:

- Shopify, Shoptet, Upgates and Eshop-rychle support product synchronization and customer order lookup.
- PrestaShop 8.2 has a **read-only order integration** through its Webservice API. Customers provide their order reference and email; order data is released only after email verification. Merchant keys grant only GET access to `orders`, `customers`, `order_states` and `currencies`. This integration does not import products or install the widget automatically. Product catalogs can be imported separately through supported product feeds, including feeds from PrestaShop, WooCommerce, Magento and other platforms.
- Knowledge sources include product feeds, website pages, sitemaps, manually entered text and uploaded PDF, Word, Excel and CSV documents.
- Embeddable chat widgets provide streaming AI conversations. Configurable sales rules can initiate conversations based on page targeting and time delays.
- Lead forms capture customer contact information. AI-assisted ticketing can summarize conversations and hand issues to the merchant's support team. CRM integrations include Zoho CRM, HubSpot and Pipedrive.
- The website and dashboard are available in English, Czech, Brazilian Portuguese, Slovak, German, Spanish, Italian and Romanian. Chat response language is configurable separately.

Developer and AI-agent access:

- The REST API is available under `/api/v1/` with authenticated, organisation-scoped endpoints for chatbots, conversations, products, manually entered knowledge base entries, tickets, leads, sales rules and analytics. Access requires an API key with the appropriate scopes.
- The remote MCP connector uses OAuth. `/mcp-readonly/` is the read-only resource for analysis workflows; `/mcp/` is the resource for explicitly authorised write workflows. Both require authentication and remain scoped to the selected organisation.
- Current MCP capabilities include chatbot and conversation inspection, product search, tickets, leads, analytics and sales rules. Authorised write capabilities include creating, updating and deleting manually entered knowledge base entries and triggering retraining. Consult the connector guide and the tools advertised by the server for the current interface.
- The MCP `search_products` tool searches the product catalog by title, description or brand. The customer-facing chat's advanced product RAG pipeline is a separate retrieval workflow.

## Core Pages

- [Homepage](https://www.chaterimo.com/en/): Product RAG, AI customer service, integrations and a live chat demonstration.
- [Product RAG and agentic search](https://www.chaterimo.com/en/#product-search): How product parametrization, hybrid and parallel retrieval, reranking and targeted follow-up lookups support product recommendations and comparisons.
- [Pricing](https://www.chaterimo.com/en/pricing/): Current subscription prices, message and storage limits, BYOK options and included provider usage. Use this page for current commercial terms.
- [AI Model Benchmark](https://www.chaterimo.com/en/ai-benchmark/): Published customer-service evaluation results and methodology, with current model rankings.
- [About](https://www.chaterimo.com/en/about/): Company background and product information.
- [Sign Up](https://www.chaterimo.com/en/signup/): Create an account, add knowledge sources, choose a model and configure the widget.
- [Privacy Policy](https://www.chaterimo.com/en/privacy-policy/): Data handling and privacy practices.
- [Data Processing Agreement](https://www.chaterimo.com/en/data-processing-agreement/): Personal data processing terms.
- [Terms of Service](https://www.chaterimo.com/en/terms-of-service/): Service terms.

## Integration Guides

- [Shopify app](https://apps.shopify.com/chaterimo): Install Chaterimo through the Shopify App Store.
- [Shoptet setup](https://www.chaterimo.com/cs/blog/shoptet-navod/): Czech installation and configuration guide.
- [Upgates setup](https://www.chaterimo.com/cs/blog/upgates-navod/): Czech installation and configuration guide.
- [Eshop-rychle setup](https://www.chaterimo.com/cs/blog/chaterimo-eshop-rychle-integrace/): Czech account connection and product synchronization guide.
- [PrestaShop setup](https://www.chaterimo.com/en/#home-integrations): Open the PrestaShop connection guide from the homepage card; install the widget and import the product feed separately.

## Developer Guides

- [MCP Connector](https://www.chaterimo.com/en/mcp/): OAuth endpoints, permissions and setup for compatible AI assistants.
- [MCP integration walkthrough](https://www.chaterimo.com/en/blog/chaterimo-mcp-server-claude-integration/): Connecting an AI assistant to the Chaterimo workspace.
- [OpenAI API key setup](https://www.chaterimo.com/en/blog/how-to-openai-api-account/): Create a provider account and API key for BYOK.
- [Anthropic API key setup](https://www.chaterimo.com/en/blog/how-to-anthropic-api-account/): Create a Claude provider account and API key.
- [Gemini API key setup](https://www.chaterimo.com/en/blog/how-to-gemini-api-account/): Create a Gemini provider account and API key.

## Languages

- [English](https://www.chaterimo.com/en/): English product information.
- [Czech / Čeština](https://www.chaterimo.com/cs/): Czech product information.
- [Brazilian Portuguese / Português do Brasil](https://www.chaterimo.com/pt-br/): Brazilian Portuguese product information.
- [Slovak / Slovenčina](https://www.chaterimo.com/sk/): Slovak product information.
- [German / Deutsch](https://www.chaterimo.com/de/): German product information.
- [Spanish / Español](https://www.chaterimo.com/es/): Spanish product information.
- [Italian / Italiano](https://www.chaterimo.com/it/): Italian product information.
- [Romanian / Română](https://www.chaterimo.com/ro/): Romanian product information.

## Optional

- [Blog and articles](https://www.chaterimo.com/en/blog/): Tutorials, integration guides, comparisons and e-commerce articles. Publication dates and historical examples may differ from current pricing or model availability.
- [Support](https://www.chaterimo.com/en/support/): Product support and contact information.
- [Contact Chaterimo](mailto:info@chaterimo.com): Product and integration enquiries.