Natural language
Customers ask normal questions without commands or knowledge of the source system.
Intelligent integrations for sales and service
OneChat AI Agents query authorized information from connected systems at conversation time—including ecommerce, ERP, invoicing, service operations and real estate inventory—to answer with context across every connected channel.

Model Context Protocol
MCP is an open standard for connecting AI applications to external data sources and tools. OneChat MCP exposes only the queries and actions enabled for each integration, while provider credentials remain protected inside the service.
Customers ask normal questions without commands or knowledge of the source system.
The Agent queries the connected source during the conversation; freshness depends on each connector.
Each connector exposes a defined set of lookups and, only where supported and enabled, actions.
The conversation can move to a teammate with its context preserved.
Seven documented legacy connectors
Commerce, ERP, real-estate and technical-service connectors have different permissions. The limits below are part of the product promise. This legacy list is preserved for technical reference; the integrations page contains the current verified catalog.
Integration for WordPress stores with expanded search, recommendations, orders, policies, tracking and notifications.
Admin GraphQL connection for products, variants, collections, inventory, orders, policies and fulfillment.
Catalog, products, SKUs, categories, pricing, inventory and orders by sales channel.
Commercial and administrative information in natural language, without requiring customers to mention Siigo.
Connect OneChat with Pony Pro by Millenium Enterprise to handle commercial requests from WhatsApp, Telegram and web chat with current information.
Current commercial information for developers and companies serving real-estate projects.
Turn service-center support into an intelligent experience with real-time lookups, traceability and automatic customer notifications.
Natural conversations
How it works
Configure the omnichannel platform, connected channels and AI Agent.
Enable compatible connectors with the required permissions and safeguards.
The Agent queries, answers, recommends and escalates according to the enabled capabilities.
Custom MCP
We evaluate its API, permissions and use case before scoping a custom MCP integration. Compatibility is confirmed during that review.