Inside Photoroom

The value of values: What we learned from an afternoon spent drawing Axolotls

Lauren Sudworth
Lauren SudworthAugust 17, 2023

What does an Axolotl have to do with Photoroom’s values? Quite a lotl as it turns out (sorry, couldn’t help myself), as we found out, when we spent an afternoon getting to grips with Photoroom’s values through the medium of animal pictures.

Photoroom's values

Photoroom’s values are a huge part of our culture: I see them showing up in almost every conversation I have across the company, from shipping a project to users in the first week, to sharing all conversations, feedback and decisions publicly in slack–DMs are actively discouraged. We have three values, and since joining Photoroom, I’ve learned them off by heart.

  • Candid feedback: We share feedback with care to challenge the status quo and keep improving what we do and how we do it.

  • Instant Learning Loops: We believe that fast iteration is the best way to learn. We're not afraid of making mistakes as there is always room for improvement.

  • Be an artisan: We value creativity and care about our product as craftswomen and craftsmen care about their work. We get joy from seeing users creating and enjoying what we build.

When you first join Photoroom, however, it can take a while to get used to working so iteratively, and having your work so regularly critiqued by other team members.

The workshop

In an effort to help make the impact of our values more explicit, I designed a workshop to express our values in a fun, creative setting (using Photoroom of course!).

The intention of the exercise was to learn how to give constructive feedback (Candid Feedback), make progress on our work quickly (Instant Learning Loops), and create something beautiful (Act Like an Artisan).

I asked my fellow Photoroomers to pick an animal they thought represented the qualities of Photoroom and draw a picture of their chosen critter. We then shared our animals in small groups, and give one sentence of feedback on what worked well, and one thing that could be improved for next time.

The rules of our critique were simple: to be kind, specific, and helpful. This framework is based on the work by the wonderful educator Ron Berger.

We then created a second iteration of our chosen animal, using the feedback from our team. After another round of feedback, we drew a final version of the creature, and then dropped it into Photoroom to create some rather magical AI/human collaborations.

What we learned

What I learned from this exercise is just how quickly feedback accelerates the learning process. Until we spelled it out, I didn’t quite realize how powerful the combination of craftsmanship, feedback, and fast, imperfect iterations can be. It was amazing what a room full of self-professed “artistic novices” (design team notwithstanding) produced in 90 minutes with just two rounds of feedback.

Even better was to see our creativity amplified by the tool that we’d all been working together over the last year. In many ways, Photoroom is a product of our collective creativity, and couldn’t exist without our values. In the fast-paced world of AI, speed, quality, and feedback are imperative, and I believe that much of our success is down to the fact that our entire team lives and breathes these qualities. Much as (you must have known I was going to shoehorn an axolotl metaphor in here somewhere), the axolotl breathes air and water.

The work

Alright, enough amphibious metaphors. Take a look at the beautiful before and after photos of the very talented Photoroom team.

Axolotl

Artist: Cato Wolff

Dog

Artist: Matthieu Rollin

Toucan (with preparatory notes)

Artist: Marc Rollin

Unicorn

Artist: Maximilien Tyc

Chimpmunk2Beaver

Artist: Antoine d'Andigné

Leopard

Artist: Jeanette Sha

Chameleon

Artist: Paulo Golovattei

Lauren Sudworth
Lauren SudworthHead of Brand and Content

Keep reading

Improving the Loading Experience in SwiftUI
Vincent Pradeilles
Core ML performance benchmark iPhone 14 (2022)
Florian Denis
Make stable diffusion up to 100% faster with Memory Efficient Attention
Matthieu Toulemont
Packaging your PyTorch project in Docker
Eliot Andres
Making stable diffusion 25% faster using TensorRT
David Briand
Building a fast cross-platform image renderer
Florian Denis
How we divided our server latency by 3 by switching from T4 GPUs to A10g
Matthieu Toulemont
The Photoroom 2023 diversity report: our pathway to a more inclusive workplace
Lyline Lim
The Hunt for Cheap Deep Learning GPUs
Eliot Andres
A picture of seven people standing in a line. One of them having just tossed a ball onto a gravel boules court.
Playing to win: the unexpected way we innovate at Photoroom
Profile Picture of Matthieu Rouif
Matthieu Rouif