Master the art of architectural thinking with AI coding assistants through real-world examples and proven techniques.
This repository contains curated sample architectures and prompting strategies designed to help developers at all levels build better software products, data platforms, pipelines, and frameworks using AI coding assistants.
Whether you're a complete beginner building your first product, launching a proof-of-concept, working on a side project, or an experienced developer looking to improve your architectural skills, this guide will help you leverage AI to make better technical decisions.
Learn how to:
- Write precise prompts that prevent AI models from making incorrect assumptions
- Include the critical architectural details that matter in your prompts
- Evaluate AI responses to identify misinformation, hallucinations, or architectural weaknesses
- Apply architectural thinking patterns to solve complex technical problems
- Start with simple, working solutions and evolve them as your needs grow
Technical architecture is like strategically filling a jar with rocks and sand.
Consider the experiment of trying to maximize the amount of rocks and sand you can fit into a jar. If you place the rocks first (representing overarching architectural decisions), the jar will hold more sand (meaning the smaller technical decisions) because there will be natural places for the sand to fill.
Source: Microsoft Copilot. (2025). AI-generated images illustrating an experiment with how to most efficiently place rocks and sand in a jar. Generated via conversation with Copilot on August 10, 2025.
Architects make high-level decisions that have long-term impact on:
- Performance: How fast and scalable your system will be
- Security: How well your system protects data and users
- Cost: How efficiently your system uses resources
- Maintainability: How easy your system is to modify and extend
Cloud architects design strategic solutions for software, platforms, and pipelines hosted on cloud vendors like AWS, Azure, and GCP. They evaluate and select cloud services by carefully weighing performance, security, and cost trade-offs to create scalable, resilient systems.
Software architects design the high-level structure of software products—mobile apps, websites, desktop applications, and frameworks. They define how software components interact and communicate, making critical decisions about patterns, technologies, and system boundaries.
Data architects design systems for collecting, storing, processing, and analyzing data. They create blueprints for data warehouses, lakes, pipelines, and analytics platforms that enable organizations to derive insights from their data.
- No prior architectural experience needed! This guide is designed for beginners
- Basic understanding of programming concepts (variables, functions, APIs)
- Access to at least one AI coding assistant (GitHub Copilot, Claude, ChatGPT, Cursor, etc.)
- Some experience using the AI coding assistant
- Clone this repository
git clone https://github.com/h-fuzzy-logic/prompt-like-an-architect.git
cd prompt-like-an-architect- Browse the sample architectures
- When you find one of interest, start with README
Learning architecture is challenging because:
- Delayed feedback: The impact of architectural decisions often isn't visible until months or years later
- Experience gap: It takes years of making (and recovering from) mistakes to develop intuition
- Limited mentorship: Few organizations have formal architecture training or continuous improvement cultures
- Isolation factor: Many developers work alone on side projects or in teams without senior architectural guidance
- Uniqueness problem: Every project has different constraints, making it difficult to find directly applicable existing architectures
- Decision paralysis: Too many technology choices without clear guidance on trade-offs
- Over-engineering: Trying to build for scale before proving the concept works
- Under-engineering: Starting too simple and hitting walls when trying to grow
- No safety net: No experienced team members to catch expensive mistakes early
- Budget constraints: Need to build fast and cheap, but don't know which corners are safe to cut
This repository aims to address these challenges by providing concrete examples, AI-assisted learning techniques, and beginner-friendly starting points that grow with your needs.
This project is licensed under the Apache License - see the LICENSE file for details.
Made with ❤️ for the lifelong learners