Elliot Gunn

If you’ve written in the past for Towards Data Science, you may also have questions about your content appearing on their new site. We’re actively working on understanding what’s happened, and if you have concerns, please email us and we’ll help educate you about your options. In the meantime, we’re getting this publication launched to serve the large number of writers who have data science stories to share, now, and want a home for them. We’ve partnered with Paolo Perrone, an editor from the Medium community, to get started (look for an introduction post soon), but we want to work with you in building out this new community-driven publication. Let us know if you want to be involved in supporting the amazing community of data scientists that has developed over the years on Medium by replying to this email and we’ll be in touch.

It's rare for companies show their hand but the Medium email re: Towards Data Science's exit is pretty revealing. Looks like it took them completely by surprise and they rushed out an email without even waiting for their new pub's intro post

It was a delight to have such candid (and funny) conversations with everyone and I'm really grateful for the generosity of the data community to share some tough stories

If you check it out, I would love to hear what resonated!

(We'll have to do a follow-up interview on CI testing!)

Not everyone is on Bluesky, so shoutout to @opendataalex.bsky.social for one of my absolute favourite interviews that's full of humour and hard-won insights:

www.datafold.com/data-migrati...

Not too long ago I posted a call for interviewees on #dataBS and I'm psyched to share what I've put together:

7 interviews with data practitioners, from AEs to DEs to architects and analysts, on their data migration stories

get.datafold.com/tl-dms

And our data migration microsite:

And that provides a good pair of rules of thumb for these kinds of statistics. Any time someone cites a very impressive quantity like this, it's a good idea to ask questions like: what's the nearest relevant comparison? What else is true if that is true (for example, if health insurance really makes $1.4tr in profit, you can Google and find that health insurers tend to report single-digit profit margins, which implies that their revenue is about equal to total GDP). You can also ask how plausible it sounds to multiply the number by ten, or divide it by ten, or if that doesn't actually seem to change anything at all.

Relevant bit for both #dataBS folks and data-inclined writers to test the sensitivity of conclusions with a pretty simple rule-of-thumb sanity check

Against Decorative Statistics If You're Going to Back Decisions With Data, Use Real Data

One of my favourite newsletters, Capital Gains, which has some really good data adjacent explainers, has a v insightful piece this week on how data literacy != better judgement.

Not all stats cited are meaningful signals - he calls them "decorative":

capitalgains.thediff.co/p/decorative...

This is really cool to read, thanks for linking this. It also gave me the idea to reach out to the authors of these types of engineering blogs, which often have a lot of technical detail but not the human side of things (decisions, trade-offs, scary moments) which is just as interesting!

DM-ing shortly!

> unrelated, just want to say that like 4 years ago I would have died at the thought of posting online (I'm not an introvert but I overthink a lot with posting anything). I feel very grateful with all the #dataBS folks helping out with reposts, sharing their stories, and DM-ing me!

Hahaha this is an awesome story (and pic) but I bet it was tough to get through at the time.

Since the statute of limitations has expired...up for a quick chat about it? Would love to dig into the technical details here

There's a very, very strong divide between lift-and-shift and rearchitect first camps. Would you be up for a quick chat on this? I would love to hear your take (+ any gnarly details)

I followed you back, it might work now! but it looks like I can DM you so I’ll do that shortly ☺️

If you're interested, it'll be a super chill, conversational interview (max 30 minutes) where I dig into the technical details + lessons learned about tools/culture/planning

Get in touch: elliot@datafold.com

Most migration stories never get shared, even though they’re full of valuable lessons:

• Unexpected challenges
• Tools that saved the day
• High-stakes moments where everything could’ve gone wrong

This is your chance to tell the story *you wish someone had told you before*

Got a question for #dataBS:

Anyone ever survived a data migration with a story to tell—the good, the bad, or the "we’ll laugh about this someday"?

I’m writing a technical piece and I’d love to hear your wins, cautionary tales, and hard-earned lessons.

Please DM if you're keen!

It’s really smart to turn an interview question into a blog post showcasing your skills. Good luck!

You could try something like: "Misaligned filters in SQL queries can lead to subtle yet critical inconsistencies in results—especially when calculating metrics like productivity over specific timeframes. In this piece, I'll...."

Grabs someone's attention immediately with what/why/how

I found this informative. One small suggestion that can have a big impact on readability: the intro needs a strong hook to answer "so what?" and this is currently buried in the middle on misaligned filters causing discrepancies in results, and what your proposed solution is

The editorial team is super friendly and their submission guidelines are straightforward.

The great thing about TDS vs your own blog is that an editor (sometimes > 1) looks over every submission and leaves helpful feedback!

The normies will follow once the power users establish themselves here lol

So we should post more about data to accelerate the migration 🤣

I think Threads does have a larger one by default but in a useless way (I have Threads only because it was easy to do as an IG user but it didn’t map to my existing Twitter follows and so I’ve never used it)

that's the thing, it sidesteps it entirely!

"but everyone is on BS" should be enough to make the decision (for folks in data anyway)

every day that BS gains followers this will become more self-evident

a politically agnostic answer would say something like, any individual decision about which app to use is inconsequential, but once the existing network starts to erode, quality starts dropping exponentially

so whichever app picks up critical mass first wins, and you should join that one

Editors’ Picks – Towards Data Science Discover the best of Towards Data Science: from hands-on tutorials and cutting-edge research to the latest on data science and machine learning tools. Your home for data science. A publication sharing...

My former employer @towardsdatascience.bsky.social has joined #dataBS!

They have a great newsletter you should check out. TDS is very welcoming to newcomers and publishes a lot, the Editor-in-Chief (Ben Huberman) does a great roundup of only the best pieces each week

Would definitely love to read this!

"The real greatest BI tool helps someone using data to orient it, so that it reflects the control levers they have to enact change."

1000% and I think this is what @cedricchin.bsky.social has been saying with his becoming data-driven series on xmr charts

tweet / databs

old habits die hard

there isn't a better label to describe what I do so I find myself saying "I'm a technical writer, but..." a lot!

"Cross-sectional" is a great way to put it. I've been struggling with how to describe what I do as my title really doesn't capture it, especially at a startup where it's common to wear many hats. I'm a technical writer that also does trade shows (sales) and creates repos (analytics engineering-lite)