Today, the current role of AI in DevOps environments is largely focused on accelerating traditional DevOps activities. However, as AI systems inevitably become more capable and sophisticated, the role of AI should be reasonably anticipated to expand dramatically. The workplace is approaching an inflection point where AI won’t simply augment middle management functions, but will assume many such functions such as monitoring tickets for completion and escalating tickets. This will fundamentally restructure how technology dependent organizations operate.
The future DevOps-centric workplace will feature a strengthened ‘Director’ role that will communicate directly with AI agents which interface directly with the ticketing and change control systems. These communications will likely be more abstract and verbal in nature, with Directors articulating high-level objectives and constraints without specifying implementation details. The AI layer will serve as an intelligent intermediary between the Director and engineering, translating strategic directives into actionable tasks.
This increased use of AI will require AI systems to have access to current policies, procedures, organizational vision documents, priorities, product roadmaps and other organizational information. AI may also need access to not just ticketing systems, but to project plans as well, so that tasks on critical paths receive appropriate treatment. Additionally, there will be the need for new forms of business documentation that will be specific to AI in support of its expanding role. These documents will not only provide operational details, but will also define guardrails to limit AI to approved tasks. These guardrails will help meet industry audit requirements (e.g. SOC2, HIPAA, ISO 27001, etc.) with regard to areas that require human involvement, such as change authorization and code review.
Perhaps most transformatively, AI will begin working around the clock within ticketing systems and project plans, actively monitoring queues and updating tickets based on pattern recognition, prioritization criteria, and personnel capabilities and availability. During what has formerly been considered downtime outside of traditional office hours, AI agents, subject to human oversight, will continuously and autonomously be completing many routine tasks such as applying and testing patches and applications. AI might even access source repositories, making code updates and flagging all such changes for human review and approval. For this level of automated change control to be practical, it will be necessary to formalize certain ‘tribal’ knowledge that typically has not historically been fully documented (e.g. task prioritization and escalation criteria). This will itself represent a significant cultural and operational undertaking.
Prior to recent AI advances, DevOps dependent organizations accumulated massive volumes of tech debt and dealt with inadequate documentation, unpatched systems, deferred refactoring and accumulated configuration drift. While elimination of some human positions may result from the ongoing introduction of AI into the workplace, it’s foreseeable that AI’s primary impact will be greater productivity and process improvement. Those professionals who wish to thrive in this post-AI landscape will be those who develop strong working knowledge of modern AI systems and an understanding of the capabilities, limitations and optimal use cases for those systems.
The future belongs not to those who resist this inevitable evolutionary change, but to those who position themselves to complement the expanding role of this key new technology. In times past it was the first ditch-diggers that realized that someone was going to be needed to operate and maintain the newly-invented steam shovel who benefited soonest and most.
