Perforce Software today added artificial intelligence (AI) capabilities to its Delphix platform for creating synthetic data used to test applications as part of an effort to make it simpler to employ a small language model (SLM) to securely test applications, even in on-premises IT environments that are not connected to the internet.
Steve Karam, principal product manager at Perforce, said rather than connecting to a large language model (LLM), Perforce has developed Delphix AI to provide access to an SLM that has been trained to use data masking techniques to create synthetic data that can be used to test applications without exposing sensitive data to DevOps teams.
Once added to the Delphix DevOps Data Platform, it then becomes possible for DevOps teams to safely test applications using synthetic data that closely resembles the data that would otherwise need to be exposed to enable application developers to build and test an application, noted Karam.
That approach also makes it possible for many organizations operating in highly-regulated industries to apply AI to testing in even air-gapped IT environments without running afoul of compliance mandates, he added. The Delphix SLM itself is based on the open-source Llama LLM project launched by Meta.
The Delphix platform from Perforce is already widely used by application testing teams to replace sensitive data with synthetic data that is realistic enough to be used to test applications that will be deployed, for example, in a specific vertical industry or geography. The AI capabilities being added to the Delphix platform make it possible to now automate that process without requiring DevOps teams to build that capability themselves, said Karam.
Next up, Perforce plans to extend this capability to also include AI discovery tools that will make it easier for DevOps teams to discover sensitive data within their application environments, he added. Additionally, Perforce is also working toward automating the delivery of data to machine learning operations (MLOps) pipelines that are used to build and train AI models.
Delphix AI is the latest in a series of Perforce Intelligence initiatives that will embed AI models and agents to better unify workflow across the company’s entire DevOps portfolio. The overall goal is to eliminate the need for scripts and frameworks to integrate workflows in favor of AI agents that share access to the same orchestration framework that will enable DevOps teams to more proactively adapt as changes are made across the entire software development lifecycle (SDLC).
As AI models and agents become more widely embedded into DevOps workflows, the next major challenge will, naturally, be adapting existing DevOps cultures to include them in a way that software engineers can reliably assign tasks to complete.
It’s not clear how pervasively AI is being applied across DevOps workflows, but the issue now isn’t so much whether it will be but rather to what degree when. In fact, as DevOps tools and platforms continue to evolve, it may only be a matter of time before many of the AI capabilities being added will soon be viewed as little more than the next set of table stakes that platform providers will need to embed if they hope to remain relevant.
