For the following projects, I had extensive dialogues with Claude, GPT, DeepSeek, Cursor, and V0 to bring my ideas to life. While past attempts with earlier models of GPT were frustrating, newer models are demonstrates remarkable improvements in understanding and bug-fixing at an astonishing pace.

My general process involves having initial conversations with the LLM (Claude, GPT, DeepSeek) to plan the functionality, structure, and layout before then creating a prototype in Figma, and finalizing the development with Cursor, V0, or VS Copilot.