Your App Just Grew a Data Center
- By Winston Thomas
- April 15, 2024
Forget fiddling with server configs by hand. That's so 2010s. Infrastructure as Code (IaC) is great, but it's spawned its own headaches.
IaC golden templates—predefined configurations or patterns used in IaC—helped for a while. Now, you can rapidly provision identical environments for testing, staging, and production without having to rewrite the IaC repeatedly.
However, golden templates were not bulletproof. They can quickly become stale as cloud services evolve, and updating them is a full-time job nobody wants.
Besides, boxing applications into existing infrastructure stifles developer creativity. It’s like squeezing the old server config into your high-school jeans and risks creating bottlenecks for innovation.
Co-pilots or AI-assisted coding do help. Developers already rely on these tools to write more code into the pipeline. But there is a hidden danger: Coding screw-ups that aren't eliminated on the fly get turbocharged.
Now, there's a stealth revolution brewing. It's where your app practically writes its own data center.
appCD: Your app, with a blueprint on the side
Sachin Aggarwal, CEO of appCD has a vision. No more handcrafted config templates and no more AI prompts that need babysitting. Picture your app, its guts laid bare, and infrastructure blueprints just appear alongside. Automation, security, and the whole DevOps wish list are baked in from the start.
appCD is part of the new generative infrastructure from code (IfC) revolution. Instead of writing IaC from scratch, generative IfC creates infrastructure in tandem with application code, leveraging knowledge from the code itself.
Aggarwal points out three immediate benefits of such an approach:
- DevSecOps alignment: Generative IfC streamlines deployment and aligns development, DevOps, and security teams from day one.
- Better standardization: Security policies are baked into the code generation process, requiring no after-the-fact checks and improving compliance. Essentially, you have a blueprint for governance at scale.
- Context-aware design: Generative IfC tools can understand the resources an application needs (permissions, sizing, etc.) from the source code, ensuring optimal infrastructure without heavy manual configuration.
This vision fueled a cool USD6 million seed round for appCD and the March 2024 early access to its software. It dissects your Python and Java, revealing the skeletal infrastructure and cloud arteries it depends on. API calls, config settings, and the whole environment sprawl out on your screen.
A new path to DevSecOps
Platform engineering teams focused on developer experience will see the most significant gains. The Generative IfC tools can also gracefully accommodate existing IaC resources, consuming templates and ensuring they adhere to best practices.
However, where appCD tools will see the biggest uptake is with DevOps teams adopting a DevSecOps mindset. Aggarwal acknowledges the inherent tension between automation and governance. “But DevOps is DevSecOps," he emphasizes, "security has to go hand in hand at that scale."
Generative IfC builds security into the pipeline, eliminating after-the-fact scanning bottlenecks. The result is a more secure posture by default.
It's also not just another attempt to create a DevSecOps blueprint. Drag, drop, lock in those connections—guardrails snap into place. Out the other end spits Terraform or Helm charts, infused from the start with the ironclad compliance of HIPAA, NIST-CSF, and the whole alphabet soup of regulations.
“Our goal is to enable any user to securely deploy applications to any cloud without delay by being application-centric, cloud-agnostic and developer-first,” stated Aggarwal in the launch press release.
Looking forward
While Generative IfC isn't a silver bullet that'll make code-prompt tools obsolete, Aggarwal's got a point. Generative AI magic needs some grounding. Sure, it can churn out code, but is it reliable, explainable infrastructure? That's a different beast.
appCD is an emerging player in this space, and its ambitious roadmap includes achieving GA by the end of 2024. Aggarwal noted that the company is already looking at helping companies with custom requirements, like having a cloud service that is not supported yet or a setting where the underlying primitive—basic data types like strings, numbers, and booleans—has not been exposed.
However, it's clear that as AI gets better at spitting out code, stuff has to change on the infrastructure side. Generative IfC might be the answer to keeping cloud-powered apps nimble, secure, and headache-free. Hype or the real deal? Watch this space.
Image credit: iStockphoto/Bogdan Malizkiy
Winston Thomas
Winston Thomas is the editor-in-chief of CDOTrends. He likes to piece together the weird and wondering tech puzzle for readers and identify groundbreaking business models led by tech while waiting for the singularity.