The Kubiya Manifesto

Manifesto: Stop Automating, Start Delegating

Prologue: The Evolution of the DevOps Role

In the dynamic landscape of technology, the DevOps movement has always been at the forefront of driving change. From the early shift of system administration to DevOps, and the rise of specializations such as Site Reliability Engineering, FinOps, DevSecOps and now Platform Engineering, we the Operators have continually adapted and shifted our attention to where the highest and best use of our talents could serve the business. However, our journey does not end by switching a job title. Today, with the rise of GenerativeAI we stand on the cusp of a meaningful technology shift that will shake the snow globe for the entire economy, and this will decide the winners and losers of business for the next 50 years.  How we prepare our organizations for this massive movement, and how well we are positioned to take advantage of this opportunity will be largely dependent on how fast we can unlearn what we’ve been prioritizing up until now and begin normalizing on a new standard.  Automation, the centerpiece of the DevOps movement, is no longer the end goal, it’s simply a means to an end. Delegation is the new Moat, and operations teams who unlock this secret and learn to leverage it to their advantage will set themselves up for winning. Where does one begin?  Here we introduce exhibit A: Kubiya AI Teammates.

‍

‍

Chapter I: The Problem with an Automation-first approach

The enduring tenets of the devops movement has always centered around the commitment to tools, people, processes. Automation which is the precipice of this standard has long been celebrated as the panacea for operational efficiency and time saving. However, while it has brought significant improvements, it is not without its limitations- we call this The Time-to-Automate Paradox.  In short, traditional automation is rigid, demanding extensive planning, coding, and ongoing maintenance - where the paradox of time-to-automation is constantly being put to the test. Standing up an internal-developer portal might sound like a great idea at first, until real world constraints set in and business ‘band-aid’ approach to shipping features faster is in inconstant competition to standardizing on a platform driven approach that may take a long time to get off the ground, if at all. More importantly, it often diverts our most talented individuals away from strategic, innovative work, trapping them in a cycle of maintaining the status quo and prioritizing low-impact activities over prioritizing high impact activities. This misalignment hinders our ability to fully leverage our greatest asset—our people.

‍

Chapter II: The new game in town Delegation- the rise of AI Teammates

Enter AI Teammates, a groundbreaking concept that harnesses the power of agentic frameworks and Generative AI (GenAI) to assume the toil and undifferentiated heavy-lifting traditionally handled by human operators. AI Teammates are not designed to replace humans but to augment and empower. They take on routine, repetitive tasks, freeing our talented engineers to focus on what truly matters: innovation and strategic growth.

Unlike traditional automation, AI Teammates are fluid, adaptable, scalable, and feel human-like. They require no exhaustive coding or maintenance and can dynamically adjust to the evolving needs of the business. 

By reducing the time-to-automation to a casual English sentence, operations teams can delegate the undifferentiated heavy-lifting roles such as solving Jira queues, allowing for self-service infrastructure, managing CICD pipelines, handling elevated permissions, and self-healing systems. AI Teammates liberate human talent, enabling them to drive forward the initiatives that differentiate and propel the organization without paying the penalty of the human talent needing to do it at the expense of business outcomes . Which Operator will complain that an AI Teammate alleviated their on-call shift in order to free their time to deliver a new environment one week earlier to host their privately hosted LLM for a new AI initiative that the company is rolling out? And where would the CIO stand on this topic? This is not a multiple choice- the answer is clear.

‍

Chapter III: The Promise of Delegation Over Automation

The promise of DevOps has always been to create a flexible and adaptable environment that aligns seamlessly with business needs and saves everyone a bunch of time. Delegation, as opposed to rigid automation, encapsulates this promise. By delegating mundane and repetitive tasks to AI Teammates, organizations can foster a more responsive and efficient operational model. This shift not only enhances operational efficiency but also empowers teams to focus on high-impact activities that drive business success, and makes their job satisfaction and career development more closely aligned to the objectives of the business.

‍

Chapter IV: The Imperative for Change

Consider this: if we do not embrace this shift, we risk becoming the next Blockbuster—an organization that failed to adapt and was subsequently overtaken by a more forward-thinking competitor in Netflix. The organizations that will emerge as winners from this change-of-guard AI Era are the ones that focus on the right priorities: being masterful resource and time allocators, and delegating away the toil of low impact activities – all the while empowering their teams to succeed by focusing the business on high impact activities.

‍

Chapter V: The Vision for the Future

Imagine a future where day-to-day technical operations are seamlessly managed (or augmented) by AI Teammates, and where your DevOps engineers step into a ‘supervisory role’ of these systems, while being free to innovate and lead transformative projects. This is not a distant dream but a tangible reality within our grasp. No CEO that we’ve spoken with takes pride in toiling over standardizing an automation stack that took years to achieve, millions to build, and is rarely in use. CEOs care about improving the business’ bottom line, and while automation is a means to achieving that goal, delegation, which is the process of removing these toil-riddled tasks from the responsibility of operations teams in favor of AI, is the END-goal.   

‍

Conclusion: Embrace the Change

What I want you all to get excited about is the work we, collectively as the DevOps movement, are doing today. If we do not begin making this shift now, or if we still are in the opinion that investing time, energy and resources on internal processes that can easily be delegated to AI,  we will find ourselves on the wrong side of history, like Blockbuster. Organizations that focus on delegating toil and empowering their teams to leverage AI Teammates will emerge as the true winners in this new era of technological advancement. That's what winning is.

Let us seize this opportunity, embrace the change, and lead our organizations to unprecedented heights. The future of Operations is here, and it is intelligent, adaptable, transformative, and natural. Together, we can redefine the landscape of technical operations and drive our businesses to success.

‍