• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

ANCHON — Learn System Design by Building Real Architectures

Inspiration

I built ANCHON because I wanted to learn system design, but I didn't want to learn it by memorizing diagrams or drawing boxes and arrows on a whiteboard.

I wanted to understand why a particular application is architected the way it is, how its components work together, what happens when something fails, and why an architecture that works for one application might fail for another.

When learning system design, it is easy to draw an architecture containing a load balancer, application servers, a database, and a cache. But drawing these components does not explain why they are needed, what trade-offs they introduce, or whether they actually solve the problem.

I wanted to move beyond knowing what components exist to understanding when, why, and how to use them.

That became the idea behind ANCHON.

What It Does

ANCHON is an AI-powered system architecture designer that aims to make system design an interactive engineering experience rather than a diagramming exercise.

Users can visually build architectures, connect components, and explore how different parts of a system fit together. With AI-assisted architecture analysis, the goal is to help developers identify potential weaknesses, understand architectural trade-offs, and reason about improvements.

For example, when designing an application, simply adding a database to a diagram is not enough. We need to ask whether the database will handle the expected traffic, whether the application needs caching, what happens if a service becomes unavailable, and whether the architecture can scale as demand grows.

ANCHON is designed to help explore these questions and connect architectural decisions to the engineering problems they are intended to solve.

The bigger vision is to make it easier to study how real-world applications are designed, understand why their architectures work, and investigate why a particular design might not work under different requirements or constraints.

How I Built It

I built ANCHON using React and TypeScript, with React Flow powering the interactive architecture canvas. Supabase and PostgreSQL support authentication and project persistence, while Row Level Security helps enforce project-level data access restrictions.

I also worked on AI-assisted architecture generation and evaluation, with the aim of making architectural feedback relevant to the system being designed rather than providing generic suggestions.

Building the project required working across the frontend, backend, database, authentication, and deployment instead of treating the architecture designer as an isolated UI.

Challenges I Faced

One of the biggest challenges was making the architecture workspace flexible and intuitive while managing complex canvas interactions, component relationships, and the surrounding interface.

Integrating the backend introduced challenges involving authentication, database access policies, environment configuration, and deployment. These problems reinforced the importance of understanding how the different layers of an application interact.

Another challenge was thinking about what meaningful architecture evaluation should look like. A diagram can be visually correct while the underlying design still has bottlenecks, single points of failure, or inappropriate technology choices. Identifying such issues requires reasoning about requirements, workload, failure scenarios, and trade-offs—not simply counting components.

What I Learned

Building ANCHON has been part of my own journey toward understanding system design from first principles.

I learned that architecture is not about choosing the most components or using the most sophisticated technologies. It is about making decisions that satisfy specific requirements while understanding their costs, limitations, and failure modes.

I also learned that a useful developer tool must do more than look good. Its interactions, persistence, security, and reliability all matter.

Most importantly, this project changed how I approach system design: instead of asking, "What boxes should I draw?", I want to ask, "What problem am I solving, why does this design solve it, and under what conditions would it stop working?"

What's Next

I want to evolve ANCHON into a platform where developers can learn system design by actively designing, evaluating, and improving architectures. I want users to be able to explore architectural decisions, understand the reasoning behind recommendations, and investigate how changing requirements affect a system.

ANCHON started as a way for me to learn system design by building something real. My goal is to turn that learning process into a tool that helps other developers do the same.

Built With

• ai
• github
• github-actions
• javascript
• llm
• openrouter
• postgresql
• react
• reactflow
• sql
• supabase
• tailwindcss
• typescript
• vercel
• vite