IBM is expected to drive AI deep into its next mainframe refresh, due this spring, in a move aimed at keeping and attracting Big Iron customers. Meanwhile, there’s no shortage of vendors vying for enterprise customers that are interested in moving off the mainframe and into the cloud. Among them is startup Mechanical Orchard, which is unveiling its generative AI-based Imogen service.
Imogen is designed to help customers move legacy workloads from their mainframes into a cloud environment where they can then be used more easily, according to the vendor.
Mechanical Orchard says its approach to moving legacy mainframe applications to the cloud is unique. Unlike some competitors, it doesn’t migrate or simply re-write legacy code “as-is” into cloud environments, according to Rob Mee, CEO of Mechanical Orchard. Instead, it reengineers the legacy systems and uses generative AI to recreate those operations as a digital twin in the cloud.
“Our approach significantly reduces the risk of failure in the modernization process, and the cost of change afterwards. Unlike many traditional methods, we focus on the behavior of the system, rather than the code itself,” Mee wrote in a blog about Imogen.
“We use a modular, iterative methodology to rewrite the living, breathing system in a modern, cloud-compatible language that’s ready for AI. Once we prove audit-level behavioral parity, we cut it over into production and decommission its legacy counterpart, incrementally and without stress. This approach radically reduces the risk of migration and gives the organization back its ability to build for the future.”
Imogen reconstructs the behavior of the system from real data flows, “rewriting components piece by piece into modern, idiomatic and innovation-ready code,” Mee wrote.
Using Gen AI in mainframe modernization
Industry research backs the ides that generative AI can address gaps in the mainframe modernization process.
According to a recent report on mainframe trends by Information Services Group (ISG), the use of gen AI can accelerate migrations and mitigate the risks associated with mainframe modernizations:
“In most mainframe migrations to the cloud, more than 70 percent of the time is spent on testing. GenAI significantly accelerates this process by generating testing scripts from the legacy source code and automatically running them on the newly created applications, ensuring speed and accuracy,” ISG’s report states. “In addition to accelerating testing, GenAI can pinpoint bugs in the code, facilitating quick fixing and improving quality. GenAI is also used as a human interface, enabling clients to inquire about the functionality of both old and new code, understand business rules and assess the impact of modifications. Other reported benefits include eliminating dead code, replacing code libraries and providing clean code for future maintenance.”
Thoughtworks partnership
In addition to the Imogen launch, Mechanical Orchard is announcing its first partner: Thoughtworks, a technology consultancy that specializes in data and legacy application modernization and cloud services software design.
The partnership will make use of Thoughtworks’ CodeConcise Legacy Assistant, an AI-based tool aimed at quickly understanding and analyzing legacy codebases. The idea is to integrate CodeConcise results into the Imogen system to speed code development, the company stated.
Mechanical Orchard was founded in 2022 by Mee, who is former CEO of Pivital Labs, which VMware bought for $2.7 billion in 2019. Mechanical Orchard has raised $74 million in funding, and it has about 100 employees. It joins a list of competitors that offer mainframe modernization services, including Kyndryl, Google Cloud, AWS, Microsoft, and Rocket Software.
For its part, IBM is readying a new family of mainframes that will be outfitted with multiple technologies aimed at making the mainframe a premier AI server.
The next iterations of IBM’s z and LinuxONE mainframes will feature the Telum II processor, Spyre AI accelerator, new operating systems, and many software additions and features designed for customers who want to deploy AI inferencing and other applications with high performance requirements.
Source:: Network World