WEBVTT

00:00:00.000 --> 00:00:05.000

- Aaron, Nate, and Jake, welcome to Talk Python To Me.

00:00:05.000 --> 00:00:07.880

- Hey, how's it going?

00:00:07.880 --> 00:00:09.160

- Hey, y'all.

00:00:09.160 --> 00:00:10.240

It's going really well.

00:00:10.240 --> 00:00:13.480

I'm excited to be on the data science side

00:00:13.480 --> 00:00:15.440

of the world today with you guys.

00:00:15.440 --> 00:00:18.040

- Cool, so are we.

00:00:18.040 --> 00:00:19.840

- Yeah, yeah, you built some really neat tools

00:00:19.840 --> 00:00:21.800

to help people get started and get up to speed

00:00:21.800 --> 00:00:26.720

and just be more efficient than just solely writing

00:00:26.720 --> 00:00:29.340

Python code, but not, you know, excluding that either

00:00:29.340 --> 00:00:30.620

with your product, Mito.

00:00:30.620 --> 00:00:33.100

So that's super fun, and we're gonna talk about that.

00:00:33.100 --> 00:00:35.820

But before we do, let's just kind of go around

00:00:35.820 --> 00:00:38.380

and how'd you get interested in data science

00:00:38.380 --> 00:00:40.260

and working on this Python tool?

00:00:40.260 --> 00:00:41.580

Aaron, you wanna go first?

00:00:41.580 --> 00:00:44.620

- Yeah, so a little background.

00:00:44.620 --> 00:00:46.740

Jake and I, you can't tell just by the first names,

00:00:46.740 --> 00:00:48.700

but we are twin brothers.

00:00:48.700 --> 00:00:52.660

So we've been working on projects together for a long time.

00:00:52.660 --> 00:00:57.580

Nate has been our best friend since middle school.

00:00:57.580 --> 00:00:59.700

I think I didn't get invited to his eighth grade birthday.

00:00:59.700 --> 00:01:01.940

So I think maybe starting in high school.

00:01:01.940 --> 00:01:04.020

And then he also went to college with us.

00:01:04.020 --> 00:01:06.260

And I think like, yeah,

00:01:06.260 --> 00:01:10.640

we got our first like taste of data science at Penn.

00:01:10.640 --> 00:01:13.180

We all studied a mix of computer science and business.

00:01:13.180 --> 00:01:14.700

And in the business classes,

00:01:14.700 --> 00:01:16.900

you do a lot of Excel based,

00:01:16.900 --> 00:01:20.180

mostly unfortunately Excel based data analytics work

00:01:20.180 --> 00:01:23.700

and stat classes and finance classes and stuff like that.

00:01:23.700 --> 00:01:25.820

So I think that's really where we got our first taste

00:01:25.820 --> 00:01:28.500

of data analytics or data science work.

00:01:28.500 --> 00:01:32.060

And then we've each had some experiences,

00:01:32.060 --> 00:01:33.780

new internships and jobs that we've had

00:01:33.780 --> 00:01:37.900

over the past few years in the data science space as well.

00:01:37.900 --> 00:01:40.220

But really, I think it all goes back

00:01:40.220 --> 00:01:44.060

to the beginnings of the courseworks that we did at Penn.

00:01:44.060 --> 00:01:45.460

- Yeah, very cool.

00:01:45.460 --> 00:01:47.820

And it's great that you all are able to stay together.

00:01:47.820 --> 00:01:48.660

I mean, obviously brothers,

00:01:48.660 --> 00:01:51.900

but continue to work together on this project.

00:01:52.940 --> 00:01:57.500

business schools, the whole business programs

00:01:57.500 --> 00:01:59.000

just run on Excel, don't they?

00:01:59.000 --> 00:02:04.180

- Yeah, it's really, it's kind of amazing, the contrast,

00:02:04.180 --> 00:02:07.060

because, so Aaron and I specifically both got degrees

00:02:07.060 --> 00:02:09.100

in computer science and in the business school,

00:02:09.100 --> 00:02:11.820

and so there was this transition where you'd be hanging out

00:02:11.820 --> 00:02:14.060

in class in the engineering school,

00:02:14.060 --> 00:02:15.140

and you'd be writing code,

00:02:15.140 --> 00:02:17.340

and then you would walk up campus into the business school,

00:02:17.340 --> 00:02:19.580

and it'd be like returning to the dark ages in some ways.

00:02:19.580 --> 00:02:20.420

- Yeah.

00:02:20.420 --> 00:02:22.420

- And what's really cool, I think, about Excel generally

00:02:22.420 --> 00:02:25.640

is that what we observed is that it let our peers in school

00:02:25.640 --> 00:02:29.280

and us as well kind of complete really amazing projects

00:02:29.280 --> 00:02:30.860

that we might have not been able to do with code

00:02:30.860 --> 00:02:33.620

because our skills, we were still learning at that point.

00:02:33.620 --> 00:02:35.440

So it's really kind of this really beginner friendly,

00:02:35.440 --> 00:02:37.480

amazingly powerful tool for what it is.

00:02:37.480 --> 00:02:39.020

But then we would go back down to the engineering school

00:02:39.020 --> 00:02:41.540

and be like, oh my God, there's all this tooling here.

00:02:41.540 --> 00:02:42.820

We could have superpowers,

00:02:42.820 --> 00:02:44.380

but we don't know how to use this stuff.

00:02:44.380 --> 00:02:46.060

And so there was this very direct contrast

00:02:46.060 --> 00:02:49.100

that we witnessed where there's very cool stuff happening

00:02:49.100 --> 00:02:51.020

all over the place, but the tooling differential

00:02:51.020 --> 00:02:53.060

is pretty dramatic between business school

00:02:53.060 --> 00:02:54.760

and computer science world.

00:02:54.760 --> 00:02:56.060

And I think that's kind of what initially

00:02:56.060 --> 00:02:57.380

made us interested in this space.

00:02:57.380 --> 00:02:59.080

- Sure, you don't want to underestimate the power

00:02:59.080 --> 00:03:02.120

of just firing up Excel, selecting a section,

00:03:02.120 --> 00:03:05.420

throwing up a graph or two, and that's incredible.

00:03:05.420 --> 00:03:09.400

And all of the functions and stuff,

00:03:09.400 --> 00:03:13.620

but if you think of bad programming practices,

00:03:13.620 --> 00:03:16.060

one of the worst ones has got to be like,

00:03:16.060 --> 00:03:18.460

do these three things and then go to over here

00:03:18.460 --> 00:03:19.540

and then do a couple of things,

00:03:19.540 --> 00:03:21.700

then go over there and then get that thing and then go back.

00:03:21.700 --> 00:03:26.380

You know, we've banished this type of programming from regular programming long ago.

00:03:26.380 --> 00:03:30.500

And Excel is like that without even being able to see where they go to his point.

00:03:30.500 --> 00:03:33.780

It's really not very predictable, right?

00:03:33.780 --> 00:03:35.020

Yeah, it's really amazing.

00:03:35.020 --> 00:03:39.420

And it's really it's crazy, I think, because once you start thinking about spreadsheets, you quickly realize,

00:03:39.420 --> 00:03:42.900

I mean, there's a couple crazy things about spreadsheets that people don't really acknowledge.

00:03:42.900 --> 00:03:46.020

We, you know, talk about them internally, because we're spreadsheet nerds at this point.

00:03:46.020 --> 00:03:48.100

And, you know, we like, you know, hyping up spreadsheets and stuff.

00:03:48.100 --> 00:03:52.540

But it's like, you know, the original killer applications of computers were spreadsheets, right?

00:03:52.540 --> 00:03:53.300

Oh, yeah.

00:03:53.300 --> 00:03:58.640

And more than that, it's, it's, spreadsheets are the most successful programming environment in the world.

00:03:58.640 --> 00:04:01.300

Hundreds of millions of people can program in Excel,

00:04:01.300 --> 00:04:04.300

the next leading programming language has 10, 20 million, you know what I mean?

00:04:04.300 --> 00:04:08.240

So it's really, there's an order of magnitude difference in how, you know, well adopted these things are.

00:04:08.240 --> 00:04:09.100

But you're totally right.

00:04:09.100 --> 00:04:14.900

The number of foot guns in Excel and the amount of like 50 megabyte insane models that we've seen where people are like,

00:04:14.900 --> 00:04:18.700

it's 75 tabs and they're all linked to each other in some crazy circular way.

00:04:18.700 --> 00:04:20.180

It's really slow. I don't know why.

00:04:20.180 --> 00:04:20.980

I know.

00:04:20.980 --> 00:04:22.580

Unclear. Not sure what's happening.

00:04:22.580 --> 00:04:25.340

But yeah, and we probably have built a couple of those ourselves.

00:04:25.340 --> 00:04:26.860

Probably even worse ones.

00:04:26.860 --> 00:04:28.580

But yeah, so, you know, we've seen that.

00:04:28.580 --> 00:04:31.700

And it's definitely those are some of the problems of spreadsheets

00:04:31.700 --> 00:04:34.540

that we kind of initially we were like, hmm, maybe these are some things

00:04:34.540 --> 00:04:36.540

that we can try to help solve.

00:04:36.540 --> 00:04:37.860

Yeah, absolutely.

00:04:37.860 --> 00:04:40.620

And, you know, Aaron, as you would go from business school

00:04:40.620 --> 00:04:44.860

back to the computer science side, I guess specifically to the business school,

00:04:44.860 --> 00:04:48.780

Did your business peers look at you like,

00:04:48.780 --> 00:04:51.300

"Oh, these are the guys that have the power

00:04:51.300 --> 00:04:52.260

to make the thing happen.

00:04:52.260 --> 00:04:54.420

They can help us build the thing

00:04:54.420 --> 00:04:57.700

that we can't quite automate or can't quite pull off."

00:04:57.700 --> 00:05:01.060

- I think in some ways, it was like trying to work

00:05:01.060 --> 00:05:03.620

in a group in an Excel spreadsheet

00:05:03.620 --> 00:05:05.340

is a miserable experience.

00:05:05.340 --> 00:05:06.740

I hope you haven't had to do it,

00:05:06.740 --> 00:05:09.660

but it's like you upload, you have a Google Drive,

00:05:09.660 --> 00:05:11.580

and then you end up uploading new versions

00:05:11.580 --> 00:05:13.980

to the Google Drive, and then it's usually a Google Drive

00:05:13.980 --> 00:05:15.780

paired with a text message group chat.

00:05:15.780 --> 00:05:18.340

And it's like, I just finished this sheet.

00:05:18.340 --> 00:05:20.540

Why don't you go up and--

00:05:20.540 --> 00:05:21.980

- Download it again or something.

00:05:21.980 --> 00:05:22.820

- Yeah, exactly.

00:05:22.820 --> 00:05:25.220

Make sure you're not doing it the same time I'm doing it.

00:05:25.220 --> 00:05:29.540

And I think that, I don't think Nate and I

00:05:29.540 --> 00:05:31.980

maybe solved those collaboration problems

00:05:31.980 --> 00:05:32.800

while we were at Penn.

00:05:32.800 --> 00:05:34.500

But I think it was those experiences

00:05:34.500 --> 00:05:37.300

and our thinking we've had of our programming

00:05:37.300 --> 00:05:40.780

as a superpower made us want to start doing this.

00:05:40.780 --> 00:05:42.820

I don't know if it was always recognized

00:05:42.820 --> 00:05:44.100

by everybody else though.

00:05:44.100 --> 00:05:47.400

I mean, I think the big thing was maybe we weren't the best

00:05:47.400 --> 00:05:48.240

at attending class.

00:05:48.240 --> 00:05:50.420

So it was hard to be a good group member in the first place,

00:05:50.420 --> 00:05:52.340

but one day we'll be up to it.

00:05:52.340 --> 00:05:54.820

- Yeah, group work was always hard for me as well.

00:05:54.820 --> 00:05:55.660

Jake, how about you?

00:05:55.660 --> 00:05:57.660

How'd you get into this whole project here?

00:05:57.660 --> 00:06:00.340

- Well, mostly by bloodline.

00:06:00.340 --> 00:06:01.700

I'm Aaron's twin brother.

00:06:01.700 --> 00:06:04.260

So sort of like some sort of covenant

00:06:04.260 --> 00:06:05.700

I think comes through that.

00:06:05.700 --> 00:06:08.420

But no, we started working really

00:06:08.420 --> 00:06:10.300

in this like Excel collaboration space.

00:06:10.300 --> 00:06:12.060

My background, I worked at a software company

00:06:12.060 --> 00:06:14.740

during college, sort of on the project management side

00:06:14.740 --> 00:06:16.100

of some data science projects.

00:06:16.100 --> 00:06:19.400

So I had not quite the business side of it yet,

00:06:19.400 --> 00:06:21.260

but at least a little more one step removed

00:06:21.260 --> 00:06:23.440

from like the coding product side of it.

00:06:23.440 --> 00:06:26.660

And yeah, we started working on these collaboration issues

00:06:26.660 --> 00:06:29.540

and we built a few other products before

00:06:29.540 --> 00:06:31.820

it had some modicums of success there,

00:06:31.820 --> 00:06:33.700

but really took a step back at one point.

00:06:33.700 --> 00:06:34.520

We're looking at like,

00:06:34.520 --> 00:06:37.100

what are the biggest problems with spreadsheets?

00:06:37.100 --> 00:06:39.220

It's the speed, it's the inability

00:06:39.220 --> 00:06:41.000

to hold onto large data sizes,

00:06:41.000 --> 00:06:43.480

and it's the lack of repeatability.

00:06:43.480 --> 00:06:46.760

This allows you to do repeatable processes

00:06:46.760 --> 00:06:47.840

in an efficient way.

00:06:47.840 --> 00:06:50.040

And so the place we found that does all those really well

00:06:50.040 --> 00:06:52.840

is some sort of bearing the lead is Python.

00:06:52.840 --> 00:06:56.420

So the idea was to stick a spreadsheet interface

00:06:56.420 --> 00:06:57.840

on top of Python.

00:06:57.840 --> 00:07:00.160

And that was sort of like a sentence we had written down.

00:07:00.160 --> 00:07:01.480

We're like, okay, now we need to backtrack

00:07:01.480 --> 00:07:03.200

and realize like, what does that mean?

00:07:03.200 --> 00:07:04.020

How do we do that?

00:07:04.020 --> 00:07:05.120

How do we implement that?

00:07:05.120 --> 00:07:07.960

And opened a few thousand cans of worms in doing so.

00:07:07.960 --> 00:07:12.160

- Yeah, I'm sure, but it's a super neat idea.

00:07:12.160 --> 00:07:15.740

There's a lot of things that you can automate with Python,

00:07:15.740 --> 00:07:18.000

but with what you guys built,

00:07:18.000 --> 00:07:18.920

we'll get to it in a little bit,

00:07:18.920 --> 00:07:22.080

but what you all's built lets you interact

00:07:22.080 --> 00:07:23.760

in this spreadsheet way,

00:07:23.760 --> 00:07:25.440

and then it writes the Python code.

00:07:25.440 --> 00:07:30.440

It doesn't just sort of allow you to make changes,

00:07:30.440 --> 00:07:32.440

and then you gotta stay in your tool, right?

00:07:32.440 --> 00:07:34.920

You use the tool to write code

00:07:34.920 --> 00:07:37.820

that otherwise might be a little bit of a stretch for you.

00:07:37.820 --> 00:07:42.700

Yeah, I think I can talk more high level, they can talk about exactly why and how we did that. But

00:07:42.700 --> 00:07:47.340

like a lot of just from the business side, like a lot of other tools will try and extract Python

00:07:47.340 --> 00:07:51.340

away. And so we'll give you will allow you to do the types of workflows that you would do in Python,

00:07:51.340 --> 00:07:57.340

but in a GUI in a visual environment, we're much more tethered to the Python or try to be more

00:07:57.340 --> 00:08:00.140

tethered to the Python and the notebook, it's really important to us that you're staying in

00:08:00.140 --> 00:08:05.180

your Python environment. And you're not at any now you're not at a disadvantage at all, because

00:08:05.180 --> 00:08:06.200

because you don't have the code.

00:08:06.200 --> 00:08:09.060

The code is right there, it's being generated in real time.

00:08:09.060 --> 00:08:10.960

And that's important for yourself,

00:08:10.960 --> 00:08:13.340

if you're learning Python, if you're trying to use the code,

00:08:13.340 --> 00:08:15.780

or if it's a communication layer.

00:08:15.780 --> 00:08:17.240

You want that code because you want to send that

00:08:17.240 --> 00:08:20.020

to a developer who's working in Python as well.

00:08:20.020 --> 00:08:22.380

- Right, yeah, it probably allows you to bring

00:08:22.380 --> 00:08:27.380

more people into the actual project than before, for sure.

00:08:27.380 --> 00:08:29.820

- Totally, yeah, you're much less silent

00:08:29.820 --> 00:08:31.380

by being in the environment.

00:08:31.380 --> 00:08:32.280

- Yeah, absolutely.

00:08:33.420 --> 00:08:40.540

I think that kind of mentality of when you're building tools for beginners or people that

00:08:40.540 --> 00:08:45.940

don't know maybe the professional software, make it really point and click and hide a

00:08:45.940 --> 00:08:47.660

lot of the complexity.

00:08:47.660 --> 00:08:50.620

I think that's something that we've experienced with tools that we've used.

00:08:50.620 --> 00:08:55.660

For example, we use Stripe and Stripe creates a bunch of dashboards for you.

00:08:55.660 --> 00:08:58.820

But the problem is we have no idea what those dashboards...

00:08:58.820 --> 00:09:01.780

What is the nitty gritty details of how those numbers are calculated?

00:09:01.780 --> 00:09:07.140

And so we have all these metrics, and we have really poor understanding of what is this

00:09:07.140 --> 00:09:08.220

actually telling us.

00:09:08.220 --> 00:09:13.860

And so I think something interesting that we definitely try to do is we give you people

00:09:13.860 --> 00:09:18.100

that are maybe less familiar with writing the syntax yourself the ability to, as you

00:09:18.100 --> 00:09:21.900

said, point and click and use the spreadsheet environment and then generate the code.

00:09:21.900 --> 00:09:26.060

And then if you ever have questions about, oh, what is this pivot table that I created?

00:09:26.060 --> 00:09:27.560

What does it really mean?

00:09:27.560 --> 00:09:31.180

Then you can look at the generated code and see exactly what's going on.

00:09:31.180 --> 00:09:35.180

And I think that kind of like understanding where users need help and where users want

00:09:35.180 --> 00:09:40.540

the as professional as possible, you know, nitty gritty details is a stratification that

00:09:40.540 --> 00:09:45.020

we've thought about and I think have a somewhat unique approach to when it comes to these

00:09:45.020 --> 00:09:46.820

like no code, low code tools.

00:09:46.820 --> 00:09:48.660

Yeah, absolutely.

00:09:48.660 --> 00:09:54.560

So I kind of want to set the stage by talking about some of the different things people

00:09:54.560 --> 00:09:59.460

are doing with notebooks because notebooks have really taken over in the data science

00:09:59.460 --> 00:10:00.980

space for good reason, I think.

00:10:00.980 --> 00:10:05.700

You know, we had iPython notebooks and we had Jupyter and we had JupyterLab,

00:10:05.700 --> 00:10:10.520

which is doing a little bit more than just Jupyter and people really love them.

00:10:10.520 --> 00:10:10.800

I think.

00:10:10.800 --> 00:10:19.140

And while JupyterLab is great, I think there's even, there's, there's a bunch

00:10:19.140 --> 00:10:23.180

of creative things going on, trying to extend it and use it in different ways.

00:10:23.180 --> 00:10:24.620

And I feel like Mido falls in there.

00:10:24.620 --> 00:10:28.140

So I wanted to throw out a couple and just see if you all have heard of these.

00:10:28.140 --> 00:10:29.820

and if so, get your thoughts on them.

00:10:29.820 --> 00:10:34.580

One of them is this thing called JUT, J-U-T,

00:10:34.580 --> 00:10:38.020

JUT maybe, something like that.

00:10:38.020 --> 00:10:40.500

And what it allows you to do is it allows you

00:10:40.500 --> 00:10:44.580

to actually view notebooks in the browser.

00:10:44.580 --> 00:10:47.500

So have you guys, or not the browser, the terminal,

00:10:47.500 --> 00:10:48.700

have you guys seen this?

00:10:48.700 --> 00:10:52.260

- I haven't seen it, but I do have a lot of sympathy

00:10:52.260 --> 00:10:54.980

for the unknown pronunciation of JUT or JUT

00:10:54.980 --> 00:10:57.060

because we get MITO and METO a lot.

00:10:57.060 --> 00:10:59.220

So my heart goes out to the two developers.

00:10:59.220 --> 00:11:02.780

- Yeah, I'm sure I'm messing it up, but yeah.

00:11:02.780 --> 00:11:04.540

Yeah, so here's a way to say like,

00:11:04.540 --> 00:11:06.800

well, these notebooks are so popular.

00:11:06.800 --> 00:11:08.680

Let's see if we can show them in the terminal.