Only 34% of enterprise teams that have evaluated existing synthetic data tools say those tools maintain referential integrity across systems, a gap Perforce is targeting with Delphix Synthetic Data, which generates scenario-specific test data through natural language prompts without sending sensitive information outside an organisation's environment.
The finding comes from Perforce's own 2026 survey of 518 enterprise technology leaders, conducted ahead of the launch. Just 36% said existing tools deliver data realism, and only 34% said they provide referential integrity, meaning the preservation of relationships between records across databases, both prerequisites for reliable enterprise testing.
Delphix Synthetic Data uses AI to scan metadata and schemas automatically, identify data structures and business relationships, and configure synthetic datasets without the manual setup that characterises older tools. Users describe the data they need in natural language and modify it through prompts; statistical analysis ensures the resulting datasets reflect the shape and distribution of source data while maintaining entity relationships across every database and file format involved.
The solution runs inside the customer's infrastructure. An embedded AI model analyses metadata only, with no customer or sensitive data transmitted to the model; organisations can also substitute their own preferred LLM running in a governed environment. Output is accessible through a user interface, REST APIs and Model Context Protocol (MCP)-enabled workflows, letting AI agents pull test data without waiting on separate provisioning pipelines.
Delphix Synthetic Data joins data masking, delivery and centralised governance in the Delphix DevOps Data Platform. Enterprise teams evaluating synthetic data solutions can request access through Perforce.
To stay across the latest in cloud, AI and enterprise tech analysis from Compare the Cloud, subscribe to our weekly newsletter at https://www.comparethecloud.net/newsletter
