Rule 4: We believe cloud infrastructure configurations have become too complex

Does your Cloud Service Provider (CSP) have a certification program? If you take a look at just one of these certifications (for example, see this handy post for one provider's “certified cloud practitioner” study guide: a 22-page document just explaining what you’ll need to know), you will find something dreaded, like this:

1. Understanding the fundamentals of contemporary continuous delivery methodologies. (Sure, seems fair)
2. Implementing and overseeing continuous delivery systems on <MyCSP> (I guess so..)
3. Establishing, monitoring, and managing systems for logging on <MyCSP>. (Wait, this isn’t built in?)
4. Skills in deploying highly-available and scalable systems on <MyCSP>. (I thought I was buying a highly-available and scalable system, no?)
5. Designing and overseeing tools that streamline operational processes.  (Did you say Designing?)
6. Implementing and automating security controls, governance processes, and compliance validation on <MyCSP>. (What? You don’t help with compliance out of the box?)
7. Twenty-one more pages of this stuff..

Navigating these complexities while keeping up with your day job - building amazing AI-infused applications - can be challenging and according to them it takes an estimated one year of hands on experience (prerequisite) to pass the “Cloud Architect” exam.  Yep, you read that right: One year, for one certification, on one cloud provider.  After completion, you will then need to get Network certified, SysOps certified, Security certified, and even a handful more to actually know it all.

Look around and you will find people proudly publishing their new Certification Badge for being “Cloud Tech Certified.”  They should be proud. It is hard work, takes real effort, and even more dedication. We applaud these professionals (and then we hire them) while at the same time, we ask ourselves this one simple question: “Shouldn’t this be easier for you, the AI Developer?”

‍

How has cloud configuration become too difficult?

Adrian Bridgwater, writing for Forbes, very clearly explained the reason cloud computing is so complex: “Cloud is complicated because computing itself is inherently complicated, and the popularization of the cloud model approach has been constitutionally riddled with chaotic platform-level mismatches that have slammed together with incongruent tectonic force.”

In short, when the pieces of a cloud stack are long and complex, the infrastructure stack itself is complex, and the distributed software that runs on it is even more complex, you end up with Complexity^3.  Worse, when every single knob, dial, switch, and lever gets exposed, you just get treated like a punching bag trying to maintain it all. Entire fields of software engineers have emerged (aka Platform Engineering) just to hide some of this complexity from DevOps Engineers (who are already highly skilled at doing this work), mainly because these trained professionals would spend all of their time running the platform versus building or writing code to run on that same platform.

‍

What about “The Cloud” is actually complex?

With such 'delightful' variables as architecture, zones, networking, storage, DNS, CPU, GPU-type and memory, cloud configuration transforms from a simple control into a riveting journey of complexity. Luckily, <MyCSP> users get to embark on the adventure of adjusting their environments and wrestling with application configurations, all in the pursuit of discovering the perfect balance between application performance and managing resource costs. Because who doesn't appreciate the intricate dance of toggling settings and deciphering cloud complexity for a touch of daily excitement?

Now let’s add billing to the mix. The total bill with <MyCSP> is an absolute breeze to manage, requiring only a keen understanding of an intricate combination of factors, including:

• How long a workload operates and on what hardware
• How much data transfers out over the internet because data in the cloud is like the Hotel California (you can never leave (without an egress fee))
• How much data a workload stores in a large number of varying ways
• Which cloud region hosts the workload as EU regions often cost more than US regions
• Whether you opt for dedicated, virtual-private, or pay-as-you-go methods
• Using managed Kubernetes services has its own unique blend of pricing intricacies and usage penalties.
• and more since we didn’t buy any software services yet

Navigating this simple billing scenario becomes even more entertaining with cost monitoring, operational monitoring, and security monitoring services, though it's worth noting that each cloud provider has these offerings too but almost exclusively cater to their native clouds. Have multiple clouds? Get used to multiple tools.

Let's face it, when entire companies are formed to help you understand or reduce your cloud bill, you know it has gotten out of hand.

What is the right solution for the AI Developer?

The AI Developer wants to bypass this pothole ridden road and find an easier path to deploying their AI-infused applications to production while focusing on TCO and regulatory concerns only.

Someone needs to deal with all that complicated infrastructure. At Seaplane we just don't think it should be the AI Developer; or a team of supporting experts they need to wait on, simply to deploy an app.

The Seaplane Platform unveils simply and operates for you a global edge network, stateless user compute, and a suite of stateful services like SQL, Object Stores, and Vector Databases; all delivered with a page of Python. Now, add all your favorite LLMs, stateless edge GPUs and a global no-cost-to-egress data network to the mix, and you've got the complete AI package. It is that simple and there are not thirty or more training sessions you have to take to get “Hello AI World” to work.  With Seaplane, write a page of Python and go to production, today.

‍

Get Access Now!

Seaplane is where AI-infused applications take flight in 2024 so join us for this ride by signing up below.  We are currently in Beta so make sure to sign up to secure your place in line and we’ll get to you as fast as we can.

‍

Sign up for the beta here!

‍