Embracing the 3-Month Rule: A Pragmatic Approach to Unscalable Solutions

In the tech world, there’s a well-known mantra from Paul Graham: “Do things that don’t scale.” However, the application of this philosophy in software development often goes overlooked. As I’ve been working on my AI podcast platform over the past eight months, I’ve adopted a straightforward methodology that I like to call the “3-Month Rule.” This system allows any unscalable solution to remain in play for three months. If it doesn’t demonstrate clear value within that time frame, it’s out.

The Challenge of Early-Stage Priorities

As software engineers, we tend to prioritize scalability from the outset. We cram our projects with complex design patterns, microservices, and distributed systems—all designed to handle massive user bases. Yet, for a startup, focusing on scalability prematurely can lead to costly delays. This framework encourages me to construct simple, albeit imperfect, code that can be deployed quickly. The hands-on experience helps me learn what my users actually require rather than what I anticipate they will need.

Current Infrastructure Hacks: My Intelligent Choices

1. Consolidated Services on a Single VM

My entire application stack—including the database, web server, background jobs, and Redis—is run on a single $40 monthly virtual machine. This setup lacks redundancy and relies on manual backups.

Here’s the brilliance of this arrangement: I’ve quickly gained insights into my true resource consumption in just two months—far more than any extensive planning document could reveal. It turns out my heavily AI-driven platform only peaks at 4GB of RAM, making a complex Kubernetes setup unnecessary and burdensome.

When the system crashes (which has happened twice), I get concrete data about real issues—often surprising ones—not what I initially speculated.

2. Hardcoded Configuration Values

Instead of using an intricate system for configuration, I have constants embedded throughout my code base, such as:

python
PRICE_TIER_1 = 9.99
PRICE_TIER_2 = 19.99
MAX_USERS = 100
AI_MODEL = "gpt-4"

No separate configuration files or environment variables. When I need to modify something, it’s as simple as redeploying the code.

The advantage here? Finding any configuration value is as quick as running a search across the codebase. Each pricing change is documented in version control, and every configuration adjustment undergoes a form of review—even