Nate Rush: Being a technical person is a mindset. It's a way of approaching problems. And to that effect, I would say that some of the craziest solutions to problems I have ever seen live in Excel spreadsheets. And really the only thing to call it is software. It's software and it's software that hides in this pretty visual interface that's been around for the past 30 years, but it's software at the end of the day. 

Eric Anderson:
This is Contributor, a podcast telling the stories behind the best open source projects and the communities that make them. I'm Eric Anderson.
Mito is a Python package that turns your data science work into a spreadsheet from a Jupyter Notebook. You can install Mito and make visual transformations on your data just like you would in Excel. All of that's encoded in Python so that you can then virtual control and schedule it. The project's taken off this year. The founders come on the show today to tell you how after a few experiments, several other years of doing other projects, Mito found product market fit, hear their growth story and more from Nate and Aaron.
We're joined today by the team from Mito, I'll sometimes call the Mito spreadsheet. This is Nate Rush and Aaron Diamond-Reivich, two of the founders of the project. Nate, Aaron, thanks for joining us.

Aaron Diamond-Reivich:
Thanks for having us.

Nate Rush:
Great to be here.

Eric Anderson:
So Mito's captured my attention, the attention of many in the community. It's a pretty exciting thing. What exactly is it?

Nate Rush:
Yeah, it's a great question. So Mito is a spreadsheet, and it's a spreadsheet that as you edit, it generates Python code that corresponds to every edit. So in practice, what that means is that analysts use Mito to edit a spreadsheet like they're used to, and in the process of that, they write a Python script that allows them to automate some reporting process that they're doing without relying on any internal engineering resources. They can do that just with the skills that they already have.

Eric Anderson:
Got it. So if they wanted a spreadsheet from beginning to end, their spreadsheets, but if they want to have one of the outputs be this repeatable Python script, then Mito's this power tool that allows them the UX they're familiar with, where they're productive and quick, and in the end, they have this repeatable process that the organization can consume.

Nate Rush:
Yeah, that's exactly right. And really, there's a bunch of reasons I think why folks choose to be in Python, which we can get into over the course of this podcast, which I'm excited to do. But in general, you have too much data to put an Excel spreadsheet for example, you might choose to be in Python, or you're looking to automate a report, you might choose to be in Python, or you decide you might want to do machine learning in the future, you might choose to be in Python. So there's a bunch of different reasons and we can get really specific about what we see in our open source community in terms of use cases. But generally, once you make the decision to be in Python, having a spreadsheet there that helps you write code effectively and use tools that you're used to, that's really where we sit.

Eric Anderson:
Sweet. And we call it a spreadsheet, but it's deployed or packaged as a Python library and you spin it up from a Python interface generally.

Nate Rush:
Yeah, it's a great question. So something that probably your podcast listeners will know is that Python open source data science, it's a big thing. And really for us, when we thought about bringing spreadsheet tooling to Python users, it was obvious like, oh, we want to exist in the context of other open source tools.
So Mito is a Python package and specifically it's an extension to a Jupyter Notebook. So it's a spreadsheet that appears inside of a Jupyter Notebook and you really fluidly can have Pandas DataFrame to use a bit of lingo, take the Pandas DataFrames, use them as a spreadsheet, go back into Pandas, et cetera, et cetera. But yes, Python package integration to Jupyter and a layer on top of Pandas as well.

Eric Anderson:
Awesome. And given now we're on the same page that we've got this Pandas explodes into a spreadsheet thing, how did we get here? What's the story behind Mito? What compelled the two of you and others maybe to pursue this?

Aaron Diamond-Reivich:
So the story goes back a little bit a ways. There's actually a third founder here, it's actually my twin brother, Jake. If he was on the podcast, we'd usually make a joke about our last names. He has a great delivery, but I'll spare you. But so the third member is Jake, and then Nate and Jake and I went to middle school, high school and college together. So we've been doing projects, doing road trips, doing other random things together for a long time. And after college, we started working on Mito right away.
I think we started working on Mito because it's a pain point, this transition from Excel to Python that we've all experienced at various points in college and at various points in our different internships throughout our college experience.
So me personally, I worked at Capital One. I was an intern at Capital One, and they had just finished this big migration into Snowflake databases, and all of a sudden, they were acquiring everybody to do their work in Jupyter and write Pandas. And coming out of business school where I learned how to do all my analytics work in Excel, I was like, oh my God, I can create a pivot table in Excel in five seconds, but it's going to take me 10 minutes to go to Stack Overflow, figure out the syntax and write it in Pandas. So that workflow is what we all got excited about supporting.

Eric Anderson:
Fantastic. And we should talk about that workflow, but also twin brother, are we identical or fraternal or should we just duplicate your voice again and then we virtually have your brother here?

Aaron Diamond-Reivich:
We're fraternal twins, but our voices people say are very similar. So sometimes if we got on a phone call, we were able to prank people.

Eric Anderson:
We could just deep fake Jake into existence here on the podcast.

Nate Rush:
It'd be relatively straightforward. Thank God they're not identical though, that'd be too much for me I think.

Eric Anderson:
So Aaron, you're at Capital One doing the business user forced into Python thing. Nate, did you have a similar experience?

Nate Rush:
I had similar transition points, but primarily in school. So I was a dual degree student in undergrad at Penn and I split my time between Wharton and the engineering school. And so for me, this transition process was a very physical one. It was like I start at the business school doing spreadsheet stuff. I'd walk down campus to the engineering school and had to become a programmer. And early days I was like, okay, spreadsheets are just better for data. I don't understand why I would ever write code. And eventually once you do make that transition, you realize all the lovely beautiful benefits that come along with not just being in a spreadsheet. But that transition for me was primarily at school.

Eric Anderson:
What I love about Mito and particularly your stories is we talk about folks who code as being technical and folks who don't code as being non-technical, but clearly there's a lot of power users and improper users of spreadsheets that can do things that coders would prefer to do in spreadsheets.

Nate Rush:
Yeah, it's really interesting. I think that this intersection space is something that we, as a company to be honest, have spent the past three or so years inhabiting. We built a bunch of tools in the intersection of spreadsheets and programming.
And I guess I'll say this, I think more than anything, being a technical person is a mindset. It's a way of approaching problems. And to that effect, I would say that some of the craziest solutions to problems I have ever seen live in Excel spreadsheets. Aaron and I have seen things that not only should no one have done, I'm upset that I've seen it. Just crazy nonsense that's implemented in Excel spreadsheets that really the only thing to call it is software. It's software and it's software that hides in this pretty visual interface that's been around for the past 30 years, but it's software at the end of the day. And I think for us, that is a really inspirational and motivational thing.
And what we think about when we think about what we're doing, we're taking those people and giving them the tools that programmers have. That's really what we're trying to do at the end of the day is let them make that transition in a much easier way.

Eric Anderson:
And in the case of Aaron, you were forced into this transition. Are people naturally wanting the best of both worlds and that's where Mito comes in? Or are most people having the two factions are at war and there's somebody who's trying to cross the divide?

Aaron Diamond-Reivich:
We definitely see both. I think our first users were people, as Nate said, they were maybe Excel first, but they're really maybe programmers at heart. And they went out wanting to learn Python because they wanted to automate their own work. Those are some of the open source users that we've worked really closely with and have given us incredible feedback and helped us iterate the tool.
We also see it from the other side where a company is trying to migrate a thousand analysts over from Excel based workflows to Python based workflows for a lot of those reasons that Nate talked about, being able to automate work, being able to version control scripts that they're creating, being able to schedule them. And so they pose different challenges in a slightly unique user profiles, but we definitely see both using Mito.

Eric Anderson:
Great. Let's pick up the story. So the two of you have this different versions of the same experience of I need to cross this divide and you all are buddies, so you go to work on this together. What's the time when you start writing code and what does that first implementation look like?

Nate Rush:
So we actually started our senior year of college, which was tail end of 2019 into the start of 2020. We started working on another different product in the intersection between spreadsheets and programming. We were trying to give GitHub style version control to Excel power users. So think of private equity analysts and associates who are collaborating on a model. Imagine if they could review and merge changes into that model, for example.
So we spent about nine months to a year working on that. That was an open source product. It still lives somewhere on GitHub. I wouldn't advise you go look, but we essentially built an open source product and realized that we weren't solving the biggest pain point in this intersection space between spreadsheets and programming tools. And we took a step back and we realized we'd been building programming tools for spreadsheet users, but actually what we wanted to be doing was building spreadsheet tools for people who were trying to be programmers. And so we'd essentially flipped it on its head, and at the end of 2020, started working on Mito.
We didn't exactly go open source immediately. So that was a bit of a process, which I think will be interesting to dive into and talk about. But 2020 is when we first wrote our first line of code, launched publicly on some Reddits and we attempted a Hacker News launch as well that didn't go so well.

Eric Anderson:
And your first foray sounds like there's VBA and there's app scripts and there's other folks who've tried to bring programming power in the context of the spreadsheet, where I think Mito's unique and I don't know if anybody's brought spreadsheets into the programming realm into the world of code. Is that fair?

Nate Rush:
Yeah, I think that's a fair phrasing. There's a lot of tools that help people write code. There's not as many people who are really specifically targeting these spreadsheet users who are making this transition. And we're unique in that sense.

Eric Anderson:
And then this first foray was open source and the second one, in Mito V1 was proprietary. Tell us about that. Was that the first one didn't work so let's try something different or?

Nate Rush:
Yeah, I think there was a bit of that. I think there was a bit of the past nine months have been such a show. Let's really flip everything on its head and start again from scratch.
Aaron, chime in here. This was first time founders coming out of school. We'd had internship experience and some work experience in school, but really we're really just getting off the ground in terms of our understanding how to build something that was really useful to people. And so yes, to be honest with you, we can look back and ascribe motivations. There was a lot of experimentation that went on just because it was like, let's try something new here. This didn't work, let's try something new. And that's really how we got our jobs.

Aaron Diamond-Reivich:
I forget, there was some company that launched. I think it was Superhuman. I became really interested in Superhuman. I signed up for the wait list and was super excited for my onboarding. So I think we mirrored some of that. When we launched Mito, we had a splash page on the website and tried to get as many people to sign up for our wait list and then do one-on-one onboarding to be engaged with the users.
So I think there was, yeah, a lot of experimentation about how to best collaborate with people.

Nate Rush:
Which to be honest, I think a lot of what we do, experimentation wise, I think reasonably could be phrased as cargo culting. To be honest, I've listened to a lot of open source podcasts. Here's how my community got off the right ground. And I think that the most useful way to think about them is ideas to add to a list and ideas to try.
Because all of this stuff is so specific to your product and your customer base and who your open source users are going to be that really the only thing to do is to try a bunch of stuff.
You can have informed hypotheses about what's going to work, but everybody's got a plan till they get punched in the face really. And for us, that's just been a repeated process over the past two years of, oh, Superhuman did this? Let's try it. Oh, that didn't work? Let's copy this other random company and eventually build up a set of tools for ourselves that really are actually effective or at least more effective, let's say. We're still learning and improving.

Eric Anderson:
Totally. So you try the Superhuman approach for Mito. How does that go and when do you switch to open source?

Nate Rush:
Yeah. Oh God. So we're in the early years of Mito. Aaron, help me recall. Because I block these out purposefully I think. So it went reasonably well. We got way more interest in Mito than we got in our initial product. We launched a very early version of the tool. So we had something that was installable. It was not open source, but it was source available so you could install it and look at the source code for it. And within our first year or so, we probably got a couple thousand downloads. We got a couple people who were interested in paying us. We experimented with maybe charging for having an open core thing or having a free version and then a paid version. So there's definitely some interest, I would say, in that thousands of downloads within the first year or so.
And then at the end of 2021, we decided, hey, let's really consider this open source stuff and see if this makes sense for where we are. That's when those conversations started.

Aaron Diamond-Reivich:
And being open source for us means a lot of things. One of the really nice things about being open source that has existed for the entirety of Mito even when we were technically closed source is the ability to integrate into open source software.
So we're huge fans of Jupyter and that's been really valuable for a ton of reasons. But one really specific one, especially in the early days was it was incredibly easy for users to get started with Mito because oftentimes the initial open source users already had Jupyter installed. They were already comfortable in that environment.
And two, Jupyter is an environment where you can write Python code. So in the beginning, we had about five features in Mito. You could maybe add a column and maybe delete a column and that was about it and maybe write two formulas. And so it was really important for us to exist in an environment where people could write Python code because if you were going to do anything with Mito, you had to also be supplementing it with your own code as well.
So we would see users and we still see users use Mito for the features that maybe now it's pivot tabling, graphing are most popular features, and then popping out of Mito, writing some custom code and then maybe putting that Pandas DataFrame back into Mito to do some additional analysis. So that back and forth between writing your own code, using Mito is a really common workflow that we see.

Eric Anderson:
Super. So I looked at before this call a little bit of your GitHub stars. You got a super popular project, it's had an incredible year, you had a jump in growth in February, one in May, one in August. And all these, I think, are at least a couple of them were correlated with Hacker News posts.
But I also noticed that you had a bunch of Hacker News posts two years ago, one year ago that didn't quite resonate as much. What's the diagnosis?

Nate Rush:
Yeah, it's a great question. There's a couple root causes that are potentially a piece of this. The first one, I think probably the most important one is the fact that early Mito literally allowed you to add a column and delete a column, which is not the world's most interesting spreadsheet. I think that's the first one.
The second thing really was going open source. So the first time we ever posted on Hacker News was two weeks into working on Mito. We essentially put up our splash page and the earliest version and said, "Hey, this is Mito. Check us out." I don't think we got a single comment outside of a few friends that we said, "Hey, can you go comment on this?" And I think we actually got a message from the admin who was, "By the way, don't tell your friends to comment." We were like, 'Oh, oops. Didn't realize. Lesson learned there." The total transparency.
But then after going open source, someone from Hacker News that we don't know and had never talked to found us and posted us. I think I got a text from Aaron that was 11:00 AM on a Tuesday that was like, "Yeah, we're on the front page." And I was like, "What?" I think I was out doing errands or something and I was furiously commenting on my phone, tiny little Hacker News on my phone, freaking out. Maybe not the best interface to be commenting on a phone, but in any case, that was totally organic. We ended up on the front page and discovered it along with the rest of the world.
And so I think probably being open source, the fact that it was linked to a GitHub Revo was probably a huge piece of the successes this year on Hacker News.

Eric Anderson:
I looked at Hacker News as the place to check, but that doesn't seem to me as the purest data science, Excel community that seems to have worked for you. Are there other communities that you go to?

Nate Rush:
Yeah, as usual, I wish Aaron's twin brother was here. He's certainly the person to talk about this, but we do a lot of work generally with data science communities. Python data science tooling is a really interesting space. What's interesting about the Python data science community is that there's a huge group of influencers and students and professionals who are interested in content about Python data science, whether they're learning it or picking up tips or just exploring new things. And so one of the things we've really tried to do is work within those communities, writing content of our own towards data science, working with some influencers as well, and educators as well to bring Mito to people and say, "Hey, this might be something that you're interested in trying. This is a tool that if you're in the beginning of your Python journey, this might be a tool that's really helpful for you in terms of getting off the ground to spend a bunch of time there."
Medium is a big one. Reddit is a big one. Twitter is a big one. Honestly, LinkedIn is a big one. Anywhere where there's influencers and social influence, some Python data scientists are hanging out as well, which is cool.

Aaron Diamond-Reivich:
I think we've also tried to get involved with Python meetups. So Jake in particular has gone to the New York City Python meetup and talked about Mito and got to meet people through that as well.

Eric Anderson:
Now that this thing's working, people are using a bunch, maybe you could tell us a bit about what adoption is like, who are these people? You've characterized who you were and who you imagined these people could be, but who's getting excited about Mito? What are the types of things they're doing?

Aaron Diamond-Reivich:
Yeah, so I think we generally think about two buckets of users. The first is probably the biggest bucket, people that are making the transition from Excel to Python. So these are maybe people that work in a bank, people that work in some financial services institution that have been using Excel for years and the company and maybe them to some degree are really excited about making the transition to Python. That's one really big bucket of people that we're focused on.
The other bucket of people that we're focused on are people that are already taking it on their own initiative to learn Pandas and Python. They maybe have taken Udemy courses or a lot of people have gone through Kaggle training programs, and they're more comfortable writing Pandas, and so they're using Mito and more augment their workflow. Creating a pivot table, no matter how good you are at Pandas, I'll challenge you and I'm sure I can make a pivot table in Mito faster than you can do it in Pandas. For some things like that, creating a graph, creating a pivot table, just being able to look at your data, being able to do that in a spreadsheet context is just vastly superior to writing out the code by hand.
So I think, yeah, those are the two probably primary user profiles that we think about.

Eric Anderson:
Looking at the first one for a moment, that seems like a big trend. If everyone in a bank who used to use spreadsheets is now having to use Python, what's going on here? I think maybe some people are like, we thought there was a bunch of quant traders and banks who were already programmers, but is this going outside of those folks? And you've alluded to some of the things that they may want, but I guess why is it happening now?

Nate Rush:
It's a really great question. So the first thing I'll say is that we definitely see this as a major trend. The first thing I'll note here, just some basic demographic data, which is depending on who you ask, Python is the most popular language other than SQL, certainly the most popular language for data science. Aaron sent an interesting article today. There's as many people today writing Pandas codes specifically, so doing data analysis in Python as there were people writing Python in 2016. Stupendous numbers and stupendous growth just in terms of who's actually using these tools.
If you look at most of these people, they're pretty much coming from two places. It's either they're coming from a data science background, usually a masters or PhD program out of education, or they're internally retrained. And so there's this huge internal retraining program where there's people with existing skills who are making this transition to Python.
And it really is, it's not quant traders. It's your analysts, a business analyst, it's a sales manager. It's people who are just day to day living in spreadsheets. And really that's the key here. If you think about this internal retraining that's happening, the most popular route is Excel to anything else. That's really the way that it goes. And that's just a function of Excel's dominance. That's a function of the fact that if you're using a spreadsheet, if you're using data in a business context, you are most likely using Excel.
Quick, crazy fact for you, if I can actually ask you to guess, if I were to ask you to guess Excel's dominance versus Google Sheets, just to put you on the spot for a moment here, what percent of the market in a business context, do you think Excel has?

Eric Anderson:
I'm probably a terrible person to answer this. So I was at Google at a time and I have a bias towards Google Sheets, so I would've hoped there'd be quite a bit more, but it's like 75% Excel.

Nate Rush:
This number is completely unconfirmed and I actually can't remember where I got it from, but we've heard 98% Excel in a business context versus 2% Google Sheets. It really is the primary route that's happening. It's Excel dominance and it's moving to new tools, and one of the primary places this is going in terms of programming language, it's Python. That's where it's ending up. And that trend is right where we sit.

Eric Anderson:
It just reveals the bubble of startup land. Because I imagine most startups are Google Sheets folks. So 98% of middle management and frontline employees are using Excel today and they're being thrust into a world of data science for machine learning for reproducibility. What's the poll?

Nate Rush:
There's a couple motivations as I alluded to earlier. I think the most acute ones are probably specific features that you can get in Python. It's usually you're looking to experiment with some machine learning pipeline. It's not usually you just not knowing any Python, you decide you're going to do machine learning. That's certainly one of the things we've seen.
Other really acute examples are you're looking to make a reproducible script. You do something once a week, you could automate this in VBA, but you'd rather not use VBA because you're not a lunatic. And so Python's a good option there.
Another big one is data size limitations. So a really acute example, we have one really fantastic open search user we've worked with a lot named Max. Hopefully that's not too much info, but he found Mito because he had a data set that was five million rows. He wanted to do one pivot table and he was like, "Oh, I can't do this. That sucks. I guess I have to find something else." And turns out, because Mito's in Python and a spreadsheet, you can make a pivot table in Mito on five million rows no problem. And so he found Mito through a Medium article, downloaded it and completed his report that day.
So those are some really specific things that we see drive people there. I'll say beyond that, there is also, as Aaron mentioned, just the person learning Python is one of the biggest buckets we see in terms of Mito users. Oftentimes people are learning Python because they hear about Python and it seems like a useful tool. It's almost part of that self tooling upgrade, a self-improvement process that they're using Python. It's not really for anything specific or acute, it's just like this seems like the language of the future, which we agree with by the way, I'm going to go learn this skill so I can market myself better, et cetera, and maybe do a better job on certain tasks that I'm doing, undefined.

Eric Anderson:
These two buckets, we maybe originally positioned them as separate trends, but maybe they're the same thing where the company is moving towards doing more Python and everyone's trying to level up their skills and self learn their way to a better job.

Nate Rush:
And I really think that that is a good unification. I think it's reasonable to say there's a bit of hype behind Python data science right now. Real businesses in some sense. There's hype in the startup world and then there's what actual businesses are excited about and oftentimes it's like, oh, we can make all of our analysts 50% more productive and automate all these reports? That sounds awesome. We're interested in Python. There's generally hype towards Python data science, I would say.

Eric Anderson:
Sweet. And then here we want to talk about the open source, but I think it's only fair to put the context around the business as well. I would think that this is a big trend to build a business on, but it's not every day you hear about a Python library becoming a big business. Pandas didn't go that route and others, but Mito seems to be poised to do so. Help us understand how that's happening.

Nate Rush:
So I think maybe the most helpful way to approach this question is talking a little about what our business actually is. What are we actually selling? So Mito is a open source project. You can download and install Mito today and you'll have all of the core functionality of a spreadsheet, able to complete your analysis and automate your reporting by generating Python code. There's also another version of the product which is effectively Mito for enterprise. So the Mito is the open core version of the product, and then we have additional high value add features on top of that that we provide to enterprises.
What we've seen, and Aaron maybe you can speak a little bit to this as well, is that enterprises specifically when you're in the context of a thousand person team that's trying to move to Python, your Python needs are very different. Usually you're also interested in trainings and certain custom types of imports from databases and oh, here's the specific PowerPoint format that we use that we really want you to put graphs in. We need PowerPoint 2006. So there's a different profile of those businesses that are making that transition. And that's really where we think about building the business and making Mito a sustainable project is that difference in what these users are looking for at the enterprise level.

Aaron Diamond-Reivich:
In terms of the enterprise offering, we're not trying to add features specifically to the enterprise offering that would be valuable to everybody and say like, "Oh, you're an open source user? Too bad, that's only an enterprise feature." It's really about what is uniquely valuable to enterprise users? So it's, as Nate said, database connections. If you're just an open source user who's using Mito in the context of Kaggle trainings or your master's program, or even if you're in a business, you would need admin support to be able to hook into your database to be able to create a dashboard that maybe you can send to other colleagues. So it's really those additional things that are only valuable in an enterprise context that we're making enterprise features.
I was also going to add to your point about Python packages that have become big businesses, I think one of our idols is Streamlet. You can PIP and sell Streamlet, and now it's part of Snowflake and crushing the world of dashboarding.

Eric Anderson:
Totally. Good. As we wind down here, you guys had a neat story around experimentation and figuring out how to get some attention and grow the community. I imagine you're still experimenting, you've figured out some things, what else are you playing with to unlock the next growth?

Nate Rush:
So we've talked a little bit about what we do, which is really working within the communities that we're trying to be useful to. I think that's a huge piece of it. There's an interesting point for us here, which is that Mito in many ways, our biggest user base are people who are making the transition to Python. They're not experienced software engineers who have this complex niche problem like a lot of open source tools are. And so as a result of that, Mito's contributor base is not 50 people. To be honest, it's a couple people opening issues and contributing maybe some to documentation as well.And so I think really the lesson there for us has been really thinking critically about who our users are and where they exist in terms of what we're trying to bring to them. That's the first place to start I think that's been the most helpful from a mindset perspective.
I think another thing that we're trying to lean into more currently is using our product itself as a tool for marketing the open source community. Because day to day, what Aaron and I actually spent our time on is not this lofty thinking about, oh, Python is cool and this is why businesses are moving to it. In practice, what Aaron and I are actually doing is talking to our users, seeing what's broken in the product, seeing what's missing and trying to add to the product, add functionality, fixed bugs, et cetera, that makes the product more useful.
One thing we don't do currently though is actually turn that around and use that to say to the open source community, "Hey, look at all these improvements we've made. This is the new version of Mito." Et cetera. So I think there's a marketing push that we could do that's really motivated by the product changes that we're making.
For example, one of the things I was working on this week is better sharing functionality, making Mito work if you're sending notebooks to each other, which is something that was a little bit broken currently, and that's a great example of something that we could turn that into a new launch, Mito 1.2. Share your notebooks, collaborate, et cetera. Tying the product into the marketing cycle a bit more is I think something that we're interested in experimenting with more as of right now.

Eric Anderson:
Yeah, I do think programmers really like the semantic versioning approach to things for obvious reasons, but at the same time, I think a lot of people have in their heads that you should just have a new version to tell me there's something new. There's a reason to look at it again and get excited about something. And semantic versioning doesn't always get you there.

Nate Rush:
Mito 3.1.1.1. Okay, nice. I don't care.

Eric Anderson:
What's next on the horizon for Mito and how can folks get involved?

Nate Rush:
Yeah, definitely check us out on GitHub. As I mentioned, we don't have a ton of outside contributors currently, but if you're interested in contributing, feel free and we'd be happy to chat about how you can get started and get set up.
The other thing I'll say here, maybe one note that's really nice to end on here, especially in the context of open source, is we really are in love with the Python data science community. So Aaron said I was on vacation last week at the start of this call, it's true. I was actually on vacation at a Jupyter conference in London hacking on some Jupyter extension stuff, and I was pretty floored by the open source community. The level of technical excellence, the collaboration that was happening across different orgs that seem like they have nothing to do with each other and why would they be collaborating? It was really heartening to me and I think it made me really excited about the Python data science space in a way that I've always been, but is not always at the forefront.
And so I think we're really thankful that this is a community that we get to be a part of. I think the open source community is exciting and motivating in ways that just building a business behind closed doors isn't, and we consider ourselves lucky that this is the space that we're building in I think.

Aaron Diamond-Reivich:
And I think one thing I'll add to is a big part of that is getting to work really closely with our open source users. So the only reason that the tool is where it is today is because of those early users that hop on Zoom calls with us once a week. It's two years later and some of the users are still getting on Zoom calls with us and giving us really specific feedback about, I'm trying to do this and I just can't. Or this button is just broken. We're super excited about listening to everybody that will use the tool and give us feedback. So if you're interested in downloading it and you check it out, join our Slack, email us, give us feedback. We're super appreciative of all of it.

Eric Anderson:
Awesome. Thanks Nate. Thanks Aaron.

Nate Rush:
Thanks Eric. Thanks for having us.

Eric Anderson:
You can subscribe to the podcast and check out our community Slack and newsletter at Contributor.fyi. If you like the show, please leave a rating and review on Apple Podcasts, Spotify, or wherever you get your podcasts. Until next time, I'm Eric Anderson and this has been Contributor.