There are a few trends I’m keeping an eye on in 2024, and I've outlined five predictions for the tech industry in the coming years. As these changes come to light they will impact how we approach cloud operations and the way teams are managed in engineering organizations. Reach out if you have any thoughts or points to add from your perspective.
1. The Rise of Specialized Clouds
Over the last few years, competition among Infrastructure as a Service (IaaS) providers has primarily centered around enhancing the breadth of services and scale they can offer. However, the next wave of disruption will be driven by specialized clouds. Instead of providing standardized cloud infrastructure, providers will increasingly compete on offering specialized infrastructure tailored to specific use cases, such as GPUs, edge compute (e.g. Akamai Cloud Computing), AI services, and frontend clouds (e.g., Netlify).
In a recent Forbes article, the prediction for 2030 was the rise of specialized clouds. It seems this future is arriving sooner, as providers are increasingly focusing on the delivery of specific cloud services. An example is CoreWeave providing GPU-accelerated cloud services. As growing companies' demand for computations continues to rise, the demand for specialized clouds is set to accelerate.
For DevOps Leaders this shift to specialized infrastructure provides a two-fold challenge: identifying the best specialized cloud service providers for their technology stack and effectively integrating these services into their existing architecture. For CTOs, this trend implies a strategic shift in choosing partners, potentially changing how they manage and negotiate service contracts.
2. GenAI-powered MLOps Reaching DevOps-level Maturity
Machine Learning Operations (MLOps) will reach the same level of maturity as traditional DevOps, driven primarily by the emergence of generative AI. The current DevOps verticals will incorporate ML versions of existing services, and new MLOps verticals will emerge. This growth in MLOps will lead to the development of specialized tools and services for model alignment, data collection and munging, model evaluation and tuning, copyright and safety checking, model observability, CI/CD for machine learning models, A/B testing, rollout strategies, and handling cloud data sovereignty of training data.
A study from Gartner predicts that more than 80% of enterprises will have used gen AI APIs and models or deployed gen AI-enabled applications in production environments by 2026, creating a surge in demand for robust MLOps practices. As ML development becomes more critical, the need for MLOps standards, tools, and practices gaining the same maturity level as traditional DevOps will be inevitable. I believe Engineering Managers will be driving the shift to integrate MLOps practices with DevOps. For CTOs, it will be crucial to keep up by building up the right team with AI talent, so workforce development and upskilling current employees will be key strategies.
3. AI Writing 80% of Code
In 2024, AI-powered tools like GitHub Copilot will be responsible for generating 80% of the code going through deployment pipelines. While this AI-generated code will significantly accelerate the development process, engineering leaders and managers will need to pay greater attention to security, reviewing the generated code, and having control over infrastructure, audit logs, and rollbacks. Leaders must ensure pipelines can manage increased code volumes and be adaptable in how code is promoted to production and rollback changes as needed.
4. 5-10x Increase in Code Volume
AI is not just helping with code volume; it's changing how teams create software. Due to the widespread adoption of AI, I predict the average codebase size will increase by 5-10x. This leads to a need for more advanced AI-powered code review and testing tools. Engineering leaders will need to focus more on process, reviews, and tooling to manage larger volumes of code effectively. Enhanced AI-powered tooling for code review, test cases, and deployment will become essential to ensure high-quality changes are implemented. New tooling generally has a learning curve and CTOs will need to ensure their teams are adept at navigating these AI-powered tools.
5. Scaling with Tools, Not Headcount
As hiring remains a challenge, CFOs will increasingly mandate engineering leaders to find solutions by adopting more sophisticated tools, rather than growing their teams. Centralized operations and common workflows will become critical in improving teams' productivity. The continued growth in platform engineering trends and fragmented cloud architectures will further drive demands for advanced developer platforms.
The 2023 KPMG global tech report highlights that 67% of tech industry leaders say that, in comparison to last year, they are expected to do more with a smaller budget. So it’s no surprise that the largest group of respondents see technology investment driving profit or performance improvement uplift of more than 10 percent, up from 2.5 percent last year. This aligns with the increasing adoption of automation tools, pointing to greater emphasis in the future on scaling with tools, not heads. I believe CTOs will steer DevOps and engineering managers towards automation and acquiring cutting-edge tools.
Annual predictions are always fun. I hope these provide insight into the changes that Zeet is anticipating in the space over the next few years. At Zeet, we are always trying to stay ahead of advancements to help our users effortlessly deploy, operate, and improve their cloud services, and you can expect us to keep pace with these changes and remain at the forefront of cloud computing.
.png)
.png)
