Contents
- 1. Automation will be the number one topic
- 2. Empowering engineers will remain hot for the next 24 months
- 3. Improving cloud cost forecasting will be the single most “AI” hyped theme of FinOps X
- 4. Expanding optimization beyond compute will garner more attention
- 5. Sustainability teams will see incremental gains in reporting by public cloud vendors
- 6. AI/ML costs will be the second question you will hear in every conversation
- 7. FOCUS – This is a game changer, and everyone will know what it is and why they need to use it by Friday, June 21st
- 8. Programming language-specific optimizations
As we are just a few days away from the much-anticipated FinOps X 2024 edition, I thought it would be interesting to share my insights on what we might expect from the participating community and vendors. Here are my top eight predictions, accompanied by a brief discussion of my anticipations from both the community and the vendors.
1. Automation will be the number one topic
We have seen in the State of FinOps 2024 survey that automation is the key driver this year and I expect to see many sessions showcasing how automation has saved organizations significant sums.
From vendors, I expect two themes: first is their integrations with common infrastructure-as-code tools such as Terraform, and second, how they have invested in simplifying their tools using automation.
2. Empowering engineers will remain hot for the next 24 months
As long as there are humans producing code, empowering them will be front and center. I expect to hear about innovative ways companies have gamified their FinOps practices and hope to see some great leaderboard examples.
From the vendors in attendance, it’s all about self-service, dashboards, and executive buy-in. Features enabling these themes will position these vendors as leaders.
3. Improving cloud cost forecasting will be the single most “AI” hyped theme of FinOps X
Accurate cloud cost forecasting remains a critical area needing improvement. I’m certain that “AI” enabled forecasting will be everywhere, however, I would caution attendees to cool their expectations. LLMs and Generative AI are not the right fit, and vendors riding on the generative AI hype-cycle will need to prove to users that their algorithms are accurate. Remember, in AI/ML, data is king.
4. Expanding optimization beyond compute will garner more attention
Optimization efforts for the past 16 years have focused on CPU and RAM. Well, with FOCUS and a plethora of tools, I expect to see organizations share their stories about how they have ventured beyond CPU/RAM and saved millions.
From vendors, I would love to see some innovations in license management, throughput analytics, deep dives into cloud services, and maybe a specialization into something like Salesforce!
5. Sustainability teams will see incremental gains in reporting by public cloud vendors
The heat wave impacting the US East Coast this week is just another example of our changing lived experience. Coupled with heightened media attention, sustainability continues to be a hot topic amongst practitioners. I expect to see renewed pressure on the public cloud providers to provide greater insights and the associated details.
Vendors can’t do much without the raw data, so I’m hoping to see some announcements that may enable everyone to level up and dig into their impact on the environment.
6. AI/ML costs will be the second question you will hear in every conversation
AI doesn’t come cheap. Be it an API, a public cloud service such as Sagemaker, a public cloud GPU, or a private AI stack in a datacenter, AI is expensive. This means that the costs associated with AI and machine learning are becoming a significant focus for FinOps practitioners. I expect to hear some great war stories and maybe even a few tactics on how attendees handled the ‘surprises’ that accompanied their AI initiatives.
Vendors will need to adopt entirely new feature sets and I expect to see some innovations in this space. Ranging from token rate monitoring with unit cost mapping through to simple enablement topics.
7. FOCUS – This is a game changer, and everyone will know what it is and why they need to use it by Friday, June 21st
Last year FinOps Open Cost and Usage Specification (FOCUS) was just getting wrapped up and attendees were waiting to see who would adopt it and how it would roll out. Fast forward 12 months and Azure, Google, and Oracle have adopted FOCUS, and I expect to hear that AWS will soon join them.
From the community, I’m looking forward to hearing how their teams have built FOCUS friendly data stores, analytics, associated dashboards, and importantly the impact on savings and budgets.
8. Programming language-specific optimizations
This may be a stretch, but I hope to hear from the community how they’ve cracked into their engineering teams’ minds and are now working on code-specific, language-specific optimizations. One hope is a solution for memory waste caused by Java.
Java is probably the most popular enterprise language, and if it weren’t for AI/ML, I’d say it’s the most expensive, but Python appears to have surpassed it. Tuning Java is difficult, especially because it lacks dynamic memory management. Without Java enabling more cloud-friendly, and possibly Kubernetes-friendly, features, we will be forced to rely on engineers and innovative vendors.
I hope to see some action in this space.
Platform9 EMP will be at FinOps X, as will I. Come say ‘hi’ at our booth, have a chat about Earned Value Management, and grab some swag.
Also read:
Tackling Kubernetes Underutilization: Cutting Costs by 50%