As my first AI-development project, I took home plenty of pearls of wisdom with this freight Saas project. Multitude of tools and technologies were used:

Honestly, coming from a design background, these terms are foreign to me and didnt mean anything. Although I relied heavily on Cursor to build this product, a lot of thinking goes into crafting the PRD to ensure the user experience is accurate logically.

Understands big code → Can read and work with long files or projects without forgetting.

Follows rules well → If you say “use Python only, no imports,” it actually does it.

Good at debugging → Spots mistakes and explains how to fix them.

Writes cleaner code → Produces code that’s easier to read and maintain.

Explains clearly → Breaks down tricky code or concepts step by step.

Works across languages → Can switch between Python, JavaScript, Java, etc.

Faster + cheaper → Runs quicker and costs less than the biggest model (Opus) but still very smart.

While Figma can be useful for conceptualising and handing off frontend designs to Cursor, my priority in this project is ensuring the backend functions smoothly with fewer bugs.

I’m confident in my frontend skills, but the backend’s intricacies are what concern me most. For that reason, I relied heavily on AI models to refine ideas and generate mockups aligned with the Product Requirement Document.

Total number of code lines:

Typical code lines for a logistics/freight platforms: 25,000-60,000