Synerise Summer Product Update 2026

Opportunities

Customer Agent knows each customer, talks in your brand voice, and sells from live product and business data, using a Persona you define. Customers ask in their own language - Agent clarifies vague requests, states its limits, and retrieves order status, stock levels, or answers from your documents.

Why Synerise:The Agent, your listings, search, recommendations and offers all run on the same behavioral AI infrastructure. What someone searches for shapes their recommendations, and what they click there changes the listings they see next. The model learns from every interaction on every surface.

See it in production: Synerise Customer Agent at Media Expert→

Opportunities

System Agent answers questions end to end using your workspace data, the User Guide, API Reference and Use Case Library. Describe what you need in plain language, like a workflow, an IQL filter or a Jinjava snippet, and it builds it once you approve. In read-only mode it runs deployment audits across the workspace, checks the status of your recommendation models and AI Search, and drafts analytical reports on how they perform.

Why Synerise:It works within your role's permissions, and nothing is saved until you approve it. Ask why a product ranked first in search and the Agent explains it. It can also tell you what the churn model predicts or how a recommendation campaign performed.

See it in action: the live demo from Synerise Way→

Opportunities

MCP Server connects any AI agent or internal tool to live Synerise data and campaigns through the open Model Context Protocol, turning natural-language intent into real platform work. No custom integration project and no exporting customer data elsewhere: integrations go up in hours rather than weeks, with OAuth authentication, permission-aware access and full audit as standard.

Why Synerise: MCP Server exposes Synerise's real object model — recommendation campaigns, Brickworks schemas, platform actions — directly to any MCP-compatible agent, secure and governed by design:

• Read-only by design
• Workspace-scoped access
• On-demand, no streaming
• Privacy-aware usage

Opportunities

Brickworks, a schema-based behavioral CMS, defines content structure once in a visual schema and delivers the whole bundle to any surface in a single API call with a zero-copy approach, grounded in first-party behavioral data and proprietary behavioral AI models.

Any field can serve different values to different segments, set once in the schema editor instead of duplicated schemas or scattered overrides. Each response can also be shaped on the fly, recommendations included, so a single call returns personalized content ready to display without extra development overhead, shortening time to market.

• Natural language setup
• Enterprise-grade validation
• Security and access control
• Version-controlled schemas
• AI governance built in
• Grounded in first-party behavioral data