My Current Project - VSKI

Last year I’ve had a lot of experiments with different technologies and learned a lot. I call 2024 “A year of R&D”. All my work in the past year laid ground for the project I am working on right now: VSKI (vi-ski) - interactive distilling system. Here I want to explain the idea behind VSKI in essence, while not leaking important details before the MVP is ready.

The essence

VSKI is a word play on the word “Whisky”. The idea is to use algorithms to “distill” the essence of the unstructured information and translate it to interactive elements. Before you ask, this project is not kind of a text generalization or ChatGPT projects, however it involves different machine learning techniques. The ultimate goal of the project is to simplify user interactions with OS and (maybe) web.

VSKI is a technology - a foudation for the operation systems without apps, it’s just another way to interact with programs. One of the main goals is to provide a platform, visual language and an extensible environment so it’ll take 6-8 month to put everything together and present an MVP and a whitepaper. In the mean time, I intend to give monthly updates on the progress if my schedule allows. Because the project’s active development just started I am sure a lot will change along the way, but anyway I am exited.

Technologies

I am not yet sure if it is going to be opensource. As of now there’s a small market for this kind of tech and I currently see only 2-3 viable use cases. So I’ll decide on the distribution model after the MPV is functional. But anyway, I want to share the technologies I’ve decided to use. No JS bro, I have enough of it at work.

  1. Flutter - I ocassionaly used it in the past and I had a team developing a flutter app, so I am sure it’s a better choice than Electron or other Web bloated tech for cross-patform UI. So the client will be in Flutter.

  2. V - I need to interact with OS and have to implement a backend. Between Dart’s FFI and some C/C++ libraries - VLang actually works great.

  3. TensorFlow - ML things

Plans

I’ve decided to publish this post because I know that some people are considering developing similar systems or maybe already in progress. So there’s no point to keep it to myself. Although I didn’t tell the actual use cases some people have context to get the picture - if you one of them DM me.

There’s no complete roadmap yet, but I aim to release a product preview in July 2025. Some demostrations and anouncements are to follow.

Peace ✌️