Submind YouTube summaries
Thumbnail for Build AI Infrastructure Skills Now or Fall Behind | Rob Hirschfeld, RackN

Build AI Infrastructure Skills Now or Fall Behind | Rob Hirschfeld, RackN

Watch on YouTube

Video summary

Many enterprises are currently adopting a pragmatic approach by equipping developer workstations with GPUs to run local open-weight models, which serves as an excellent entry point for exploring AI capabilities. However, this strategy is not sufficient for long-term enterprise success because it lacks the necessary control and auditability required when deploying systems across individual end-user machines. Relying solely on decentralized hardware limits visibility into what users are doing with their resources, making it difficult to govern approved models or ensure compliance. In contrast, a dedicated cluster of AI resources offers superior cost controls, governance capabilities, and the flexibility to manage diverse workloads effectively without being restricted by the specific configurations of individual desktops. Furthermore, the hardware landscape for artificial intelligence is evolving rapidly, with new chips emerging that provide inference capabilities even on standard CPUs, meaning organizations should not lock themselves into GPU-only solutions found only in personal devices. Beyond hardware considerations, enterprises must also plan for agentic systems and batch operations that require a controlled environment rather than relying on ad-hoc setups like cracked-lid laptops scattered across an office. Guaranteeing service level agreements (SLAs) is critical to prevent workflow disruptions when AI services go down or perform poorly, ensuring that routine processes can continue effectively using open-weight models without risking operational stability. The path from initial experimentation to owning robust AI infrastructure involves building skills in multiple dimensions simultaneously rather than following a rigid, sequential order often seen in traditional migrations like VMware projects. While empowering developers with local tools is an immediate first step, it must happen alongside efforts by hardware and operations teams to learn how to set up Kubernetes for AI workloads, select appropriate models, and integrate new capabilities into the broader infrastructure. Companies that focus on building these intermediate skills early end up innovating faster because they do not wait until a future date to address operational control, automation, and process improvements; instead, they embed these competencies across their organization right from the start. Ultimately, successful AI adoption requires a holistic strategy where delivery of infrastructure, model selection, and system integration occur in parallel rather than as isolated phases. Organizations should avoid getting tied up solely in deciding which specific large language models to use or how to configure individual harnesses before establishing the foundational skills needed to build scalable systems. By ensuring that all teams are developing the necessary expertise concurrently, enterprises can create a flexible infrastructure capable of adapting quickly to rapid AI innovation while maintaining strict control over their assets and data. This proactive approach ensures that businesses remain competitive and secure as they transition from local pilots to comprehensive, enterprise-grade AI operations.
Read the full video transcript
Many enterprises are starting by putting GPUs in developer workstations and running local open weight models, which is a great way to get started. But why is that a reasonable place to begin, but not a complete long-term strategy? >> Well, there's a couple of reasons why not and and one of them is innovation and pace of innovation. So, I want to I want to get there and and put a pin on that and how you do it. But from an enterprise perspective, ultimately, it's going to be about control and auditability. So, what we're looking at here is if you're running an enterprise system and you want to have your AI um you know, your your end your individual end users, you said developers, but it could be any end user leveraging a local model, that is going to save you on processing capabilities. However, what you're going to do is be limited to the what that end user is doing on their machine, when their machine is running, when you have access to it. And it's going to be very hard to audit and control and check. So, the idea of having a cluster of AI resources actually gives you better cost controls, but it also gives you better auditability and governance, lets you manage those pieces. And this is where the other the other dimension comes in. Inference is does not necessarily require a GPU. There's a lot of new chips coming coming into market that will provide AI inference capability. And you know, even running CPUs and CPUs are getting new inference capabilities. So, the you know, locking yourself into what's on people's desktops, it's a good short-term solution, but long-term, having that that that dedicated AI capability with a heterogeneous mix of hardware, so you can match models to capabilities and bring in new new models whenever you need to and then control to make sure that the models that you've approved are getting used, those control points are absolutely essential and really from an enterprise perspective critical to scale. And it's it's worth noting when you look at how those pieces go. And that's exactly what we're helping companies spec and build today. The harness allows you to to pull those pieces in. You guarantee an SLA, right? If you've been We We get caught up in AI goes down, has a day where it's not performing well, and a lot of work gets stopped. Um you're going to want to be able to control even if you're not on a frontier model. Having guaranteed access to models to run processes is absolutely essential because more routine processes can be be run very effectively by open weights models. And along those lines, what you also need to think about is in that infrastructure and with those controls. And there's one more critical point with this, which is that it's not just your users, your developers, and people doing this. Part of what you need to be thinking about building is agentic systems, and those agentic systems have to run somewhere also. And you don't want a uh room full of cracked-lid laptops providing your agents. You actually want them running in controlled corporate infrastructure. And so, all of these things work together to drive this this realization that you want to have a dedicated AI infrastructure that can run your inference models in a controlled way, allow your agents to work, especially offline or or to provide, you know, batch operations in a controlled way. So, all of these pieces fit together. So, as much as I love running local models and empowering developers, from an enterprise perspective, you need to be planning further out. >> And now, what does the path from that first step to owned AI infrastructure actually looks like for an enterprise? >> Well, this is one of those ones and actually it's funny cuz VMware migration is similar to this in our books and experience. Is enterprises have a tendency to try and sequence out all these steps and they they think they have to do this in a very orderly manner. Our experience actually is that the path for should have a lot of parallel operations. You know immediately, right, that you can empower developers to do local models and look at Harness. So, Harness is is clearly a first step, but the people who are going to make an evaluation for your Harness, right, your platform team, your developers, aren't the same ones who actually are the ones who should be looking at how to run and inspect AI gear and infrastructure. That's mostly hardware teams and operation teams that need to understand how to set up and run AI gear, how to build Kubernetes from an AI workload perspective. And what I would what I would encourage, especially because AI innovation is running so quickly, is that you need to be building these skills in multiple dimensions simultaneously. So, you should be looking on how do I improve my delivery of AI infrastructure? How do I know what to buy? How do I know how to wire it together? How do I build these systems and get that skill set embedded in your organization. Uh we see exactly the same thing going on with a lot of VMware choices where people try to figure out what their exit out of VMware should look like and then work backwards. The reality is the all answers require you to have better operational control of your infrastructure, more automation, better processes. This is what RackN does from a bare metal perspective. What we've seen is the companies who build the platform and the capabilities that allow them to then innovate on top of that platform end up moving a lot faster than the ones who figure out where they have to get to at the end of the trip and don't worry about, you know, any of the intermediate skills they're going to need to build. Especially in today's market, you need to be making sure you're jumping through all the intermediate skills simultaneously to build this up. So, don't don't get tied up in I need to figure out my harness. I have to figure out which LM is best. I have to You do need to do those things, but don't wait on I'm going to be building AI infrastructure until after you've made those decisions. Start building all the skills across your teams so that they're delivering the pieces that you need right out of the gate.