Data Scientist

## The work

We make financial research videos.

Half the job is code: finding data, scraping it, cleaning it, and structuring it.

The other half is figuring out what the data actually says — finding the insight, building the story, and communicating it clearly through charts.

You’ll work directly in a shared GitHub repo with the rest of the team.

## What you'll do

* Use AI and coding agents to research, scrape, and structure data
* Pull data from SEC filings, FRED, APIs, and the web
* Clean and organize messy financial data
* Build charts in code
* Find interesting economic and financial trends
* Turn those findings into clear visual stories
* Commit your work directly into our shared GitHub repo
* Figure things out without needing every step explained

## What we need

* Experience using AI and coding agents to build real things
* Comfortable working directly in GitHub
* Real scraping experience
* Comfortable working with messy data
* Able to build charts in code
* Strong analytical thinking
* High agency
* Clear written English

SQL, JavaScript, financial knowledge, and strong design taste are bonuses.

We can teach you finance.

## To apply

Start your response by writing **Purple Donkey**.
Please write the AI models you used when doing the test. (We expect you to use AI.)

Then complete this test:

Find an interesting economic or financial insight using data from **FRED, SEC filings, or both**.

Build a short visual story around it using at least **3 charts in sequence**.

The charts should work together:

**Chart 1:** establish what is happening
**Chart 2:** explain it or add context
**Chart 3:** reveal the implication or conclusion

Pull and process the data with code.

Submit:

* Your code
* The underlying dataset
* A PDF containing the charts and a brief description for each chart

We care about two things:

**Can you get the data right?**

**Can you turn it into a story worth telling?**

Great information design is a bonus.

We skip applications that don't follow the instructions.