Glassbox launches AI tools for enterprise digital data

Glassbox has launched Glassbox Pulse and the Glassbox MCP Server, aimed at bringing digital experience data into enterprise AI systems.

The tools are designed to turn customer behaviour, technical issues and business impact into operational intelligence that can be used in real time. Glassbox is targeting sectors where digital failures can quickly affect revenue, trust and service levels, including banking and insurance.

AI context

Pulse is an AI-powered engine that identifies changes in digital journeys and highlights issues that need attention. Glassbox says it uses machine learning to classify behavioural and technical signals, then passes relevant information to teams or AI systems instead of exposing them to large volumes of raw journey data.

The Glassbox MCP Server uses the Model Context Protocol to make that information available inside enterprise AI environments. This allows AI applications and agents to access governed context and supporting evidence without ingesting a continuous stream of underlying customer journey data.

The launch reflects a wider shift among software suppliers seeking to position themselves as intermediaries between operational systems and generative AI tools. Rather than calling for ever-larger data flows into models, vendors are increasingly promoting filtered, structured and auditable information as a way to keep AI outputs relevant and easier to govern.

Industry focus

According to Glassbox, Pulse can detect problems such as stalled applications, failed payments, release-related friction, chatbot issues and technical errors. Teams can also ask questions in natural language and receive responses linked to evidence including customer behaviour, journey impact, errors, network logs and server-side failures.

In regulated industries, the emphasis on evidence and auditability is likely to be central to the sales pitch. Financial institutions and insurers often face tighter scrutiny over automated decision-making, customer treatment and the handling of sensitive data. That has increased demand for systems that can show why an issue occurred and who was affected.

Guy Perry, Chief Executive Officer, Glassbox, outlined the company's view of AI's role in operations.

"Enterprise AI has to move from answers to action," said Guy Perry, Chief Executive Officer, Glassbox. "Glassbox gives organizations the trusted context to resolve issues faster, protect customer trust and drive measurable digital outcomes at scale."

Perry said companies also need to be more selective about the data they pass into AI tools.

"Our customers don't need more noise; they need an auditable dataset that binds customer behavior to its technical cause, proving exactly how errors tie to conversion drops and lost revenue," said Perry.

Standards push

Glassbox focuses on digital experience intelligence across web and mobile channels, analysing how customers move through online journeys and where friction appears. The company says seven of the 10 largest US banks by assets are among its customers and that it serves organisations in more than 40 countries across financial services, insurance, retail and other digital-first sectors.

The introduction of an MCP Server also points to the growing influence of common protocols in enterprise AI integration. As companies deploy more AI agents and assistant-style applications, suppliers are looking for standard ways to expose internal data sources and operational signals while maintaining controls over access, relevance and privacy.

For Glassbox, the challenge is to show that digital experience monitoring can move beyond retrospective analytics into live operational decision support. Its argument is that if AI systems receive filtered evidence tied to customer journeys, technical causes and business impact, they may be better placed to help teams respond before abandoned journeys, service calls or lost sales begin to build.

Glassbox says its platform holds ISO 42001 certification for AI management and ISO/IEC 27701:2019 certification for privacy management.