Mindbox has integrated the Model Context Protocol (MCP) to enable AI agents to manage marketing tasks within its ecosystem. These agents can now assist in creating email campaigns, automated scenarios, and marketing creatives based on text descriptions or visual inputs like screenshots and brand guidelines.

The AI agents are capable of performing complex tasks such as drafting newsletters for both desktop and mobile, inserting product recommendations, and auditing existing campaigns for configuration errors. According to product manager Irina Belitsa, this replaces the manual assembly of campaign blocks.

The system provides access to over 50 data marts, including CRM revenue, website funnels, and A/B test results. Users can query this data using natural language, prompting the AI to generate SQL queries and provide analytical insights without the need for manual report compilation.

Security measures ensure that personal customer data remains within Mindbox's distributed infrastructure in Russian data centers. AI agents operate exclusively with aggregated metrics and project settings, and they lack the permissions to delete campaigns or initiate the final launch of communications.

Early testing has shown significant efficiency gains; for instance, the company Garlyn utilized the MCP integration to launch 20 automated scenarios and 40 mass mailings within three days. Future updates will expand these capabilities to include pop-ups, SMS, and push notifications.