Software engineering performance significantly impacts the building of sturdy, stable applications. The community aims to imbibe the engineering principles typical of software development, including systematic approaches to design, testing, development, and maintenance. This requires the careful amalgamation of applications and metrics to warrant complete control, awareness, and accuracy.
This could be achieved through practices such as infrastructure as code (IaC) for deployments, automated testing, app observability, and complete app lifecycle ownership. The DevOps Research and Assessment (DORA) team discovered four key metrics crucial for gauging the performance of a software development team: frequency of deployment, lead time for changes, change failure rate, and time to restore service. While these metrics measure the efficiency and effectiveness of DevOps practices, the focus is often on distributed technologies and the cloud, leaving mainframes in a unique position.
BMC Software integrated AWS Generative AI into its product BMC AMI zAdviser Enterprise. This uses Amazon Bedrock to offer analysis, summarization, and improvement recommendations based on DORA metrics data. However, tracking DORA 4 metrics could lead to individuals feeling scrutinized, prompting a needed shift in organizational culture. It’s crucial to avoid focusing on irrelevant metrics and excessive data tracking to maintain the essence of DORA metrics.
BMC AMI zAdviser Enterprise enables proactive identification and resolution of issues by providing DevOps KPIs to optimize mainframe development. The solution pairs API calls to Amazon Bedrock with zAdviser Enterprise to provide summarization, analysis, and recommendations for improvement based on DORA metrics KPI data.
As part of the solution, no personally identifiable information (PII) is contained in the API call, nor does the generative AI client retain, learn from, or cache this data. The implementation also significantly reduces the barrier to building AI-driven organizations. Large language models (LLMs) can offer huge value to enterprises wanting to explore and use unstructured data.
With continuous collection of organization-specific DevOps metrics, Amazon Bedrock APIs equipped with BMC’s zAdviser’s industry-specific knowledge demonstrate the power of AI in the field. This generates potential for a series of tasks to be performed, beyond chatbots, including classification, editing, and summarization. With this approach, BMC zAdviser Enterprise was integrated with Generative AI, highlighting the transformative strength of the API.
