VSKI Update - Changing Directions?
Getting more insights almost always means that the selected course will be adjusted or changed completely. Long story short - this is exactly what happend to my side projects. And I think this change is for the better.
For the past month my focus was set mostly on networking and indirect market research. Indirect approach in my case means that the research is mostly about gathering information on issues through the talks “on the side”. This is actually a hard endevour because usually people won’t talk about it outside their workplaces for obviuos reasons. But nevertheless I was able to scrape some important bits and cross-reference it with my own experience. Thanks to Malaga’s and Valencia’s awesome IT communities and environment.
Why VSKI isn’t public (yet)
There are various reasons I keep this project description vauge. First and most important is the “patent trolls”. You know those bots downloading your public repos on every update? It might be a Joe you took a gig from 7 years ago, and if you do something similar they might consider it their IP. I learned it from the others mistakes - I won’t ellaborate, just google it. Some of the “Joes” didn’t know that we track referers, so I know at least what “compliance software” they use :). This is why my public repos are only PoCs and code examples.
The second reason is trivial - I don’t have as much resources as some of the competitors, so going public and spending resources on an undercooked project is a big risk. So while we’re cooking only a small circle knows the details.
Changing directions
Originally I saw only a couple of use cases for VSKI as an ML meta-framework and tried to develop something flashy to demostrate the use cases, but after getting some perspective I came to the conclusion that I would invest as much time on creating something useful for industry. The storage engine in this case is the best way to go. So without revealing the important details I can tell that the product itself will be an object data storage.
As it turned out most of the businesses I know accross industries struggle with similar issues, and it turns out that some of the issues could be solved with the techniques I originaly developed for VSKI, and some of them I solved numerous times in different contexts. Thanks to the insights I’ve got from the recent meetups with awesome business owners and engineers (I’ll write about those event in the future posts).
After everything is ready, the first lauch will be through the accelerators, and also - I’ll directly contact some businesses I have worked with. I’ve decided to try a hybrid distribution model - Binaries (shareware), SaaS, and Tailored Services. I took some pages from FerretDB team experience because I like how they entered the market. VSKI won’t be opensource, at least on first stages, but it doesn’t mean the business wouldn’t be able to get access to the code.
I caught myself spending in a 2 week limbo moving buttons instead of developing the important parts - most of the devs know how it is. So I completely ditched the idea to develop flashy UI at the first stages, as it is not so important. There will be minimal desktop and web, but I don’t see any reason creating a flashy UX because it won’t add any value to the product, so I’ll focus on the speed and other optimizations.
Timelines & Jurisdictions
The timelines didn’t chage and may be even accelerated. What chaged is where. Currently I am in EU but I don’t consider any EU or Asian countries startup-friendly. Too much unnecessary regulations, high taxes for small returns, slow decidion making, etc. France and Germany are mostly OK but not w\o flaws. And in current politics the markets may change. So I’ll be picky with the accelerators, but I definetely want my endevour to be global rather than tied to a restrictive jurisdiction.
If nothing changes in US laws in following year, perhaps we’ll go with US (Delaware) because it is still a good choice for non-residents with an access to global and US martkets. But it might also change.
Building a team
I have engineers I trust who are ready to jump on board when there is a client and problems to solve. So to build a mobile team isn’t a problem for me. However if it comes to support and scaling the solution to consumer market then HR risks would be a concern. This is why I talked about “Binary” distributions and SaaS. If my endevour is successful I want to maintain a mobile, professional team instead of scaling omni-directionaly. In fact, I believe that the best product are built by the small teams. Take Telegram or Signal for example.
Sum up
- The product is a storage engine (database) with nieche features (currently undisclosed)
- The lauch dates have not changed (Q4 2025)
- First, the product will be offered to the clients I’m in direct contact with, then to the accelerators
- The team will be as small as possible. We’ll need engineers.
Peace ✌️