Building What I Keep Reaching For
Ever wish you could more easily switch between Claude, ChatGPT, Gemini, Nano Banana Pro 2, etc., depending on your task? Each of the models is the best at certain tasks. With these models, once I've had the best at a particular task I have to have it every time for that task until the model is dethroned.
I'll spare you the explanation of the use case here. If you're at my site, reading this far, you understand.
So I've started building an app that's missing in the AI space. I don't know what the name for this would be, but it's:
- an internet browser
- an AI chat browser
- an AI voice assistant
- project mission control
- a beginner-friendly coding assistant
- a pro-level coding assistant
- an audience builder
- a marketer
- a marketplace
...with backend agents running 24/7 via cloud and local models, and world-class memory via a hybrid of .md file management and RAG vector DB configurations. I am using the SDK kits from Anthropic, OpenAI, and Google Gen AI.
Three goals, memorialized
I want to memorialize three specific goals to re-visit them later for accountability. The app must:
- Be easy to use for a beginner.
- Give the user an automatic, proactive personal AI agent experience with zero upfront configuration. Out-of-the-box, ready-set-go.
- Allow and facilitate users of all levels to have a simple mechanism for sharing workflows in a form that is plug-and-play to other users, while benefiting both the creator/poster of the workflow (with more than stars or likes), and in turn providing a benefit to the broader community.