Episode Transcript
Kyle James (00:00.965)
Hey, welcome to the AI Chronics podcast. I'm your host, Kyle James. And today we're going be diving in headfirst into how a software company called Rose is using AI inside of their own business. And we'll share the exact steps that you can take in order to implement AI for yourself. Now, before I dive into that, listen closely. Are you looking to implement AI inside of your own company? Or maybe just struggling to get your AI to stop hallucinating? Speak to GPT Trainer. GPT Trainer literally builds out and manages your AI for you.
eliminating hallucinations for good. Go to gpt-trainer.com. I promise you it'll be the biggest time saving decision that you've made all year. Trying to set up AI on your own is like trying to build a house from scratch. Sure, you could do it, but the time and frustration is going to take you to get it finished just isn't worth it. It's a thousand times faster and safer to hire professionals. Schedule a consultation today. Once again, that's gpt-trainer.com.
Say hi with me, Torben Schultz, who is the co-founder of rose.com, an AI spreadsheet for data analysis and reporting. Super excited to have him on the show today. Hey, Torben, welcome in.
Torben (01:10.976)
Okay, thanks for having me.
Kyle James (01:13.157)
Yeah, so tell us a little bit now, like what is Rose? And mean, obviously you're using AI, but like, how did you find the company? What exactly does it do? So some of our listeners can understand what it does.
Torben (01:24.408)
Yeah, sure. Well, the initial idea of Rose came when my co-founder and I, we were building our last company and we had also been former management consultants. So we've been kind strong spreadsheet users in our past lives. And we thought, okay, well, Excel, was actually pre-ChatGPT that we came to the idea and we thought, well, Microsoft Excel, 40 year old technology.
Google Sheets as the major innovation with online collaboration founded or created in 2006 before the iPhone was invented. So it's like fairly old technology that like one to 2 billion users around the world are using. And our initial hypothesis was that we wanted to make the data import much better because everyone works with APIs and with fast tools now. And we wanted to make the UI better, more notion-like, prettier.
more kind of flexible, you can use it as a dashboard and so on. And that's what we built initially. And then when the chat JTP came, came upon us, we were kind of the first movers in the spreadsheet world and jumped on this opportunity. And now AI has become our biggest growth driver.
Kyle James (02:42.929)
Yeah, absolutely. like in that process, like when you first found the company a couple of years back, was AI in the midst? Like was it already embedded into the system ahead or did you add it on later on to the company?
Torben (02:58.028)
Yeah, we first actually set out to build the foundations of a spreadsheet. Turns out that building a full Microsoft Excel from scratch takes a bit of time and a bit of engineering work. So that's what we actually did first. And then we built the AI on top of it. So with ROSE, you have a full spreadsheet platform. You can actually, if you want to, write any
Microsoft Excel-like function like VLOOKUP, some, some if, and so on yourself. You can use pivot tables and charts and you have a fully fledged spreadsheet platform. But now with AI, it's gotten way easier to use it. You can be much more productive in your data analysis and reporting.
Kyle James (03:44.613)
Yeah, so walk us through that a little bit. like you added AI, obviously built the framework for it, you know, having all the functionality, but then you added AI. Like what exactly is the AI doing? And like, walk me through that kind of step by step on that process.
Torben (03:58.73)
Yeah, so our users mostly use AI for data analysis and for reporting, including like data transformations for data cleanup and so on. And well, our products are actually available. You don't even need to sign up for it. You can just go on rose.com or ws.com and sign and well try it out immediately.
You would start by adding some data. You can use one of our integrations or you can upload a CSV or Excel file. And then you're ready to use the AI analyst. it's kind of, well, now we're in the world of Vibe coding. So you might call it Vibe spreadsheeting or so. So you have a side panel that you can use for prompting.
And then you can ask either vague questions like, well, give me some insights about this data set, or you already have something concrete in mind and you want to actually build something like step by step, like for example, well, clean up the date column, right? Or add a column with buckets of revenue by 1 million or so, right? Or add this type of chart.
or add a pivot table that does this or that. We also have a Python integration, so you can also ask for more advanced data analysis like build me code model and it will for something that takes me as a pretty solid spreadsheet user like a couple of hours, it will do it in a matter of 10 or 20 seconds.
Kyle James (05:41.179)
Yeah. So there, I imagine there's probably a broad range of different types of users with, you know, who use rose.com and, like for specifically, and feel free to, to elaborate on any, any, like any different parts of, the business, but like what types of results have you been seeing so far from maybe some client feedback that they've been giving to you or even internally seeing some feedback? Like what's something that's worth sharing?
Torben (06:06.274)
Yeah, think, I mean, what we saw in the last few months was when well, in the last few months, we did a couple of things. we, we launched the new generation of our AI analyst, we added the Python integration to it. And we also made it available on mobile. So on on you, you can actually go to roast.com on your cell phone, and you'll have kind of a
an interface that you can switch from spreadsheet to copilot or AI analyst and back and forth with toggle. And that can toppled our usage of the AI analyst. It kind of tripled in a matter of weeks the number of customers that we had on the AI analyst. So we used to, well,
Our growth driver used to be integrations, but now it's clearly become the AI. Internally, think what we see and what people really like is we also have a smart data import. For example, if you have invoices in PDFs, you can just use our upload feature. It doesn't only work with CSVs or Excel files, it also works with
PDFs and image files. And then you can actually use a prompt for the import. you can say, well, extract the supplier name, invoice number, invoice date, the amounts, tax amount, gross amount from these invoices and combine everything into a table. And it will do that really nicely. Even the invoice structure can be completely different. So from different suppliers and so on.
and it will do it super nicely. And that's a feature that people love in particular.
Kyle James (08:06.791)
It's almost like kind of like a sandbox here where you're like, instead of having to go through manually, okay, this is I'm going to change is just having the kind of like the AI, you tell it what you need and it automatically changes exactly what you need to on an invoice. So they're having to go through manually or having to send it back to another team member, have them change it. Like you can change it manually yourselves, which I imagine that's probably a huge time saver for a lot of people who are using it, using rows.
Torben (08:29.964)
Yeah, exactly. Especially for like small businesses, know, smaller finance teams and so on where where you're getting kind of a ton of invoices from different suppliers, usually they end up in some email inbox or so and then and then parsing through them or importing them and so on. It's it gets quite time consuming in a lot of ways and this is way we're using AI to parse it.
is actually quite nice. We also, if you go on our social media channels, we actually just open source the part of this and kind of published a white paper about it if you're interested in it technically.
Kyle James (09:13.167)
Yeah, yeah, for sure. And maybe as we, you know, obviously there's so much change happening within the AI space and you know, I love how you mentioned that, that one number where he's like, Hey, you added the, think you mentioned you added the AI and tripled, like tripled having the amount of users, which is incredible. That's, that's incredible to hear and going into that, knowing that kind of data on the backend, what kind of upcoming AI initiatives is, is Rose planning out over the next year or two? And where do you see it playing, maybe becoming some of the biggest role in your operations next?
Torben (09:43.522)
Yeah, I think our vision for the AI analyst is really like, you tell it what you want to do with your data and it does it just like your data analyst or your business analyst in your company would do, right? And delivers you that type of output. so obviously people do
very many different things with spreadsheets. So there are lots of different contexts that we have to adapt it to. well, one big difference to, for example, doing that in chat GPT. mean, people also use chat GPT to kind of upload a CSV and then ask something about it, right? The big difference here is that it's not kind of a throw away analysis. You will get your report in an interface that you can keep and automate.
and share with others and so on. Right. So that is the big advantage. What we're planning to do now is well, one part of it is making the data imports much smarter. So we have over 50 native integrations, but most of them we can't use yet through prompting. So there's kind of still manual work to do. Let's say you want to get data from
Stripe or Google Analytics or so, right? Or BigQuery. And then enabling this completely through prompting, let's say we're a management team and we're using BigQuery or Postgres or something and we have our integrations connected and we are sitting around the table and asking a business question about, how did sales develop in the last weeks and so on, rather than...
someone running out of the door and then we have to wait for 30 minutes until the analysis comes in. We can just prompt it, ask the question and get the answer as a report immediately. That is kind of the vision that we're pursuing there. A second part I was saying where people throw lots of different analyses or different questions at a spreadsheet. I think currently we're really good if you
Torben (11:58.636)
have, if you go step by step, it's like a way more productive spreadsheet building. I think we're currently working on improving kind of the answers to vague questions, right? That we give you like a much better answer, kind of a fully fledged report to a vague question.
Kyle James (12:21.671)
Yeah. Yeah. I, when you mentioned the example of like sitting down, especially like in a boardroom meeting and seeing like all the C-suite, all the head leaders and like, Hey, well, how did sales go? Hey, how did revenue go? Hey, what were some of the questions? What were the top? And it's like, uh, we didn't put that on our presentation, but in this case, if you had some like that built out, then now you're talking about where you can just ask it live, ask the AI kind of the AI analysts live and give you that hard data right there.
on a presentation within some of the C-suite, which I would imagine that's probably for a lot of people out there listening, they're like, gosh, I can't tell you how stressful I've been on some of those presentations and not having the answer for the founder or the CEO. But in this case, if I had an AI that can give me that answer quickly, I would probably worth.
Torben (13:09.71)
Yeah, and if you're in this situation, like in the best case, it takes 10 minutes. In the worst case, it takes maybe a week or never, right? So yeah, I think that is something that really makes decision making much faster, right? And companies way more product.
Kyle James (13:16.379)
Hahaha
Kyle James (13:27.515)
Yeah, 100%. So as we wrap up here, Torben, appreciate having you on the show today. And where can people learn a little bit more about you and maybe a little bit more about Rose that you would recommend them checking out?
Torben (13:39.054)
Sure, mean, you can just go to our website, rows.com, R-O-W-S, rows like columns, and start using the product right away. And you can then, if you like it, sign up for a free plan where we're freemium. And if you use the AI or the integrations more, then you can upgrade to one of our paid plans. We're also on LinkedIn and on Twitter and on YouTube. So, yeah.
feel free to check out our social media channels here.
Kyle James (14:13.007)
Awesome. Love it so much. Torben, it's great having you on the show, brother. We'll definitely keep in touch. Hopefully, we may be happy on the show again next couple of months. Who knows, right? And thanks for everybody for listening in today. And remember, if you're looking to implement AI into your business today, don't try and do it yourself. The time and stress that the AI could cause may not be worth it. Schedule a call with GPT Trainer and let them build out and manage your AI for you. Once again, that's gpt-trainer.com. Signing off for now.
Have a great rest of your day, everybody. Looking forward to seeing everyone on the next episode of AI Chronicles.
