Stitch Fix, an online clothing styling company, built an interactive page showing how its algorithms and data science work.
Commenters praise it as a fancy, accessible explanation of its recommendation systems and cite it as a model for presenting algorithm write-ups.
> Thanks! Really appreciate it! Yeah eventually we'd like to provide a writeup of the data and algorithm that is accessible - something like: https://algorithms-tour.stitchfix.com/
Whoa never saw that before, fancy as. Yeah, I feel that a nice writeup would lend credibility to your algos.
> Did you find the scoring and guided experience helpful or would you rather just a big list of criteria you can filter from? We played around with the former but think the latter may actually be more appealing to more people. That way compatibility is binary, either it matches your criteria or not.
You see this is where it gets tricky. The people that hang around HN tend to be more technical and likely used to more convoluted UIs that allow them to be efficient and flexible.
At the same time your market is likely to be less techie so I fear that my opinion may be biased and even harmful.
That being said I supposed having a clean and simple default with an advanced alternative a click away would satisfy both crowds though the latter can wait post-launch so you can see what the feedback is like.
As a quick note I think even the simple version would benefit from 'justification' behind the recommendation, i.e., we suggested Hawaii because while expensive it has top-notch surfing (assuming they selected that), is safe, LGBTQ-friendly, etc. Even if you skip the numbers you make it easy for people to compromise on things and rethink their choices before trying again. Otherwise I feel it'd…
Stitch Fix | San Francisco, CA | Full Time | ONSITE ONLY | General/Backend Software Engineering | http://algorithms-tour.stitchfix.com
Algorithm Development Platform Team seeks software engineers to help data products get built at Stitch Fix.
We're looking for solid software engineers that like to be their own product managers, love ambiguity and setting their own priorities!
You'll get to architect and build "Lego Bricks" that Data Scientists use to build data products.
Experiences wanted:
* Building apps, libraries, tooling from scratch. Backend service experience required.
* Seen apps at scale, as well as infrastructure for making service development and deployment quick and easy.
* Background/interest in ML is a plus.
So if you're interested, please apply via:
* Please apply via https://www.stitchfix.com/careers?gh_jid=852996&gh_jid=852996 or
* Cover Letter encouraged -- please mention hacker news!
The Interview experience is:
- phone chat with me (the hiring mgr & tech. lead)
- phone screen (1 or 2)
- maybe a take home (piloting this right now)
- onsite (2 half or 1 full)
Me: http://www.linkedin.com/in/skrawczyk
Stitch Fix | Fullstack Engineer | ONSITE (SF) | Full-time
We're looking for someone who is a capable fullstack engineer.
The Algorithms team at Stitch Fix has developed tons of interesting and powerful models over the past 6 years. Our team is looking for someone to help us prototyping many new product ideas that combine our algorithmically capabilities in useful ways for our clients. This role is very much an exploratory product development engineering role with the end goal of developing a new product to bring to market.
This role requires that you can quickly develop high quality products that demonstrate novel, useful interactions between our customers and our algorithms.
You'll likely be a good fit if your skillset overlaps with:
- React, Swift, Python, Next.js, Node.js, Tailwind.css, GraphQL, Algorithms -
You can learn more about the algorithms we've built at https://algorithms-tour.stitchfix.com
If you're interested, please reach out at tmillar@stitchfix.com
Thanks! Really appreciate it! Yeah eventually we'd like to provide a writeup of the data and algorithm that is accessible - something like: https://algorithms-tour.stitchfix.com/
Did you find the scoring and guided experience helpful or would you rather just a big list of criteria you can filter from? We played around with the former but think the latter may actually be more appealing to more people. That way compatibility is binary, either it matches your criteria or not.
Rome2Rio are definitely a company we'd like to learn from and partner with. There's a surprising amount of aussie travel startups (adioso, wotif etc) that have done well.
Hi! Some background first: I'm putting together a blog right now using Hugo and D3. I'm a huge fan of D3's infinite flexibility, as seen in some famous scrollytellers [0-1], and I've spent some time experimenting with that format myself [2].
My question is: what does Observable Framework offer for data storytellers who want to blog? Is this meant to go up against Hugo/Jekyll in terms of full-fledged max-efficiency site generation? If not, are there plans to add integrations with other blogging frameworks?
[0]: http://r2d3.us/ [1]: https://algorithms-tour.stitchfix.com/ [2]: https://vermarish.github.io/big-brother-barometer/
Twenty points for him linking to a site (his non-brutalist stitch-fix link) that gave me an aneurysm after trying to scroll for a few pages. Not sure what he's trying to say there. "I learned how not to respect my user"?
( https://algorithms-tour.stitchfix.com/ , should you also want to punish your cortex ).