Getting a real-time picture in rentals is hard without the right system: what’s out, what’s due back, who’s using it, and whether an order is fully served. Rental complexity explodes when every item is unique and time matters. Think of it as a 3D “data cube”: unique items × attributes × time. Availability and utilization optimization are basically cube queries—computationally heavy without the right model.
- Why linear retail’s SKU counts don’t work for rentals
- How item-level attributes + temporality create a “data cube”
- What “situational awareness” really means in rental operations
- Why you need an operating system (not just an app) for workflows and APIs
- Where TWICE aims to be: Linux-grade backbone with Apple-level UX, built for AI agents
Tuomo: Getting a situational picture of a what's happening right now for a rental business, say, vape resorts and rental business there, can be quite challenging if you don't have the appropriate software for it. So what's out, what items are out, what's coming in, what are we supposed to give out today, where are all of the people who is using them, what's the state of these orders, like commercially have we served now everyone in this one order that has been paid. So all of that becomes quite complex and similar aspects then happen in all of the re-commerce side of things. Yeah and then maybe we'll probably get to a bit to this same topic when we start to talk about connectivity, but to mention earlier we were talking about temporality, every item being unique and then this temporality in also in terms of orders. So I like to think about sometimes, I like to think about it as like a data cubicle and I'll explain what that means but if I look at the linear commerce model, I might have like a 2D plane, I have a 10,000 SKUs and then for each one of those I have a number and that's what I track and then that number is going up and down, but I don't too much care about the individual item that is representing one of those increments in the number. Now the first thing that happens if when I go single SKU, let's say that I on average per SKU I have thousand items, so that 10,000 because 10 million data rows and now each one of those have all of the columns that are defining its attributes and the taxonomy of it, the condition and the color and the product information, the monetary information, all of that becomes trackable and on a unique level. Now we connect temporality to it, which is kind of again a new axis, so now we have a 3D cubicle and it goes from all of the time that these items were created to infinite future, so then your cubicle starts to grow to be quite big. This is kind of the cubicle inside which your situational picture as a business lives and then your business questions start to be what's the state of this cubicle next year, January 5th and then you need to go into that cubicle, find the relevant stuff, filter out what is interesting and computationally that becomes quite complex and then things like availability optimization, utilization optimization is effectively looking at that data cubicle and figuring out what is the best item for example to serve and order. For me if I think about it through that lens the complexity becomes understandable but still like you respect the data cubicle.
Karri: Absolutely, I think if there is this huge data cubicle but at the same time you have to be super flexible to the individual needs of each business then I guess that's a good time to not talk about software but more like actually operating system where you are enabling this.
Tuomo: Yeah, it's like when you start to design these solutions it's not designing just like a vertical software, it's more of like the mindset if it's building a operational system, operating system so it's like trying to understand what defines this operating system, why is it great to use via I don't know comment line or via graphical user interface or via just API calls or similar. So it goes to this level and unlike as a design problem but of course usually the greatest operating systems have some underlying assumptions that define what can be done in terms of computation and such but then they also have maybe a graphical user interface for that for users and then that's the world of preferences many tend to like for example Apple's quite simple user-friendly approach but then there are the people that want the granularity and that that other OSS might be providing.
Karri: And where does twice fit on this scale if you are talking about simplicity Apple and then Linux Unix type of...
Tuomo: I think we want to build the Linux I think already being kind of coming from Finland you know Linux Turvals it's like I think building Linux would be probably the dream when it comes to the graphical user interface I think we'd like to be in the in the Apple category at one day as an approach because well in the in the age of like AI and AI agents and AI augment the UX what's the role of UI in five years of time I don't know it's it's it might be that we can spin out quite custom UIs on the fly almost so maybe the question is then more in the like Linux level stuff and how accessible is it for maybe AI agents via I don't know MCP layers and all of that as twice we are maybe navigating towards that Linux and maybe Apple hopefully.
Karri: Sounds like a pretty interesting topic for another podcast episode.