Embracing the 3-Month Rule: A Pragmatic Approach to Unscalable Solutions
In the world of startups, the mantra “do things that don’t scale,” often attributed to Paul Graham, is frequently cited but rarely unpacked, especially when it comes to coding practices. After dedicating the last eight months to developing an AI podcasting platform, I’ve discovered a practical framework that I call the 3-Month Rule. This approach helps me evaluate the viability of unscalable solutions within a defined timeframe.
A New Perspective on Scaling
As engineers, we are typically conditioned to think in terms of scalable solutions right from the outset. We immerse ourselves in design patterns, microservices, and distributed systems, focusing on building architectures capable of supporting millions of users. However, this mindset can lead to what I like to call “expensive procrastination” at the startup stage—spending resources optimizing for users who aren’t even on the platform yet.
My 3-Month Rule encourages me to embrace simplicity in my coding efforts, allowing me to produce direct, albeit rudimentary, code that is deployable and insightful. Here’s a look at some unconventional (yet effective) infrastructure strategies I’ve employed during this journey:
My Current Infrastructure Hacks
1. Consolidated Operations on a Single VM
I run the database, web server, background jobs, and caching on one $40/month virtual machine (VM) with no redundancy and manual backups.
Why is this approach effective? In just two months, I’ve gained significant insight into my true resource requirements. My AI-driven platform has surprisingly low memory needs, peaking at only 4GB RAM. The excessive Kubernetes setup I almost pursued would have been managing containers that were never utilized.
When outages occurred, I obtained valuable data on what truly failed, which has consistently been unexpected.
2. Hardcoded Configuration Values
My codebase is filled with constants such as:
plaintext
PRICE_TIER_1 = 9.99
PRICE_TIER_2 = 19.99
MAX_USERS = 100
AI_MODEL = "gpt-4"
With no separate configuration files or environment variables, updates require simple redeployments.
This approach allows me to easily track every price change through Git history and ensures thorough review of each config update. Constructing a dedicated configuration service would have consumed a week of my time, yet I’ve modified these constants only three times in three
