Composable Infrastructure with Adjustable AI: How Clockwork Lets You Control the Automation

How Clockwork Works: Composable and Adjustable

Imagine you are building a house. Most of the time, you want to build each room yourself, but sometimes, you might want help with the tricky parts. Clockwork is like having a smart assistant who can help as much or as little as you want. You can say, “Just handle the plumbing” or “Design the entire kitchen, but let me place the cabinets.” It is the same with infrastructure. You can write every detail yourself, or let the AI fill in the gaps based on simple descriptions. For example, you might say, “I need a web server that can handle a lot of traffic,” and the AI will choose the best software and settings for you. But if you care about which specific web server to use, you can specify that too. This flexibility makes it powerful yet simple.

The key innovation here is the adjustable AI. Think of it like a car with an automatic transmission that also lets you shift manually when you want. You can start in fully automatic mode, and if you need more control, you shift gears manually. Similarly, with Clockwork, you can start with a high-level description and let the AI handle the details, but if you want to tweak something, you can take over without starting from scratch. This is especially useful for developers who want to avoid repetitive tasks but still keep control over critical decisions. For instance, you might use AI to configure a database but then manually adjust the storage settings because your application is particularly write-heavy. This balance is what makes Clockwork unique.

• Specify only what you care about, and AI handles the rest
• Same resource type can be used with different levels of control
• Group resources together for coordinated AI configuration
• Assertions let you validate without locking implementation

Why This Matters: Beyond the Hype of AI

Under the hood, Clockwork uses Pydantic models to define what you want. You might create a DockerResource and say, “This is a web server,” and the AI will figure out it should use Nginx or Apache, configure the ports, and set up logging. But if you care about which web server it is, you can specify that too. You can also group resources together, like saying, “I need a development environment with a database, a cache, and a web server.” The AI will not only choose the right tools but also configure them to work together, for example, setting up the database to connect to the cache automatically. This composability means you can build complex systems from simple, reusable parts.

This approach is not just about convenience. It is about making powerful technology accessible. When you can start with a simple description and then refine as needed, you are more likely to try new things. If you had to configure every detail upfront, you might stick to what you know. But with Clockwork, you can start with, “I need a machine that can handle 1000 requests per second,” and then later add, “And it needs to be secure,” or “And it should send errors to a monitoring service.” This iterative process is how complex systems are built in the real world. You start with a goal, and you adjust as you learn more. Clockwork brings that same iterative, exploratory approach to infrastructure.