Indie Log #7

I've started a new project called Gennmix. I know Writeflow was my main project, but I believe this one is more interesting. So what is Gennmix?

Gennmix

This era is becoming filled with new AI models. Every morning, I receive updates about newly launched products, and it's all amazing.

Lately, I've been incredibly excited about generative AI models for images and videos. It's as if a whole new world of possibilities has opened up. I've finally admitted to myself that I should use AI as much as possible to keep track of these developments.

I believe Cursor is the most valuable product for me. Although Cursor doesn't have its own AI models, it leverages external APIs like GPT and Claude effectively. Similarly, I wanted to aim to create a Cursor-like tool for content creators. This is the core idea behind Gennmix.

I've already written and deployed some code, and even promoted Gennmix on Reddit. However, the current feature set is limited. At present, Gennmix is essentially a simple web interface that allows users to generate images through various AI services using their own API keys.

The generated images aren't stored on the server. Instead, their URLs are saved in the browser's local storage. This approach, however, creates a vulnerability where users might lose their images if their local storage is cleared or compromised.

I plan to enable image storage for logged-in users. Before begin this work, I need to implement various AI services for image generation, editing, and background removal. Fortunately, I've written code that quickly builds forms using only Zod schema, which will make this process more manageable.

Future Plans

The final form of this app is still uncertain. My primary goal is to build rapidly and gather user feedback. I intend to develop the core features as quickly as possible, share the project publicly, and create a Twitter account to promote it.

By the way, I mentioned earlier that I had already posted about this project. I shared it in the DALL-E 2 channel. Unfortunately, it didn't generate any interest. I'm trying to determine if there's actually a need for this product, which is challenging for me.