Charlotte de Thiersant

In case you've been wondering why your users were not assigned properly to your feature flags/experiments anymore, @Amplitude_HQ made a major change: they stopped creating flag and experiment user properties ({flag-key} = assigned_variant) 🤯 community.amplitude.com/discussion/169…

@photoroom_ML is now connected to Shopify!! 💜💚 Now, you can manage your entire listing catalog in a single place! For anyone running a Shopify store, this is going to save you real time and real money. Try it now and let us know what you think 👉 lnkd.in/eRznDrDg

@guillaumesimon @photoroom_ML Love thissss

Colleague : I wish i had a tool to... Me : Turn on Tv to watch 🏉 , Launch Claude code, Use the skill i created to use @photoroom_ML design system, watch France crush it, deploy to vercel before half time. Voila.

@Valentin_Fgrs @iMart1n @evanspiegel sameeee hahah

@iMart1n @evanspiegel Thanks Martin, I feel so rich now!!

ThinkDifferent™ @evanspiegel

@criteo @zenly @photoroom_ML 45 minutes pour démystifier la data produit : 👉 Créer une vraie culture data, au-delà des simples metrics. 👉 Définir la bonne durée pour un A/B test 👉 Choisir les bons outils pour faire scaler une "Modern Data Stack".

@criteo @zenly @photoroom_ML Je suis ravie d'avoir pu en discuter au micro de Killian Sowa dans son podcast "On t'parle app" youtube.com/watch?v=VZdQ_J… linktr.ee/otpa

J'ai eu la chance de travailler dans trois boîtes "data-driven" qui sont de vraies réussites françaises : @criteo , @zenly et @photoroom_ML . Leur point commun ? la data est au cœur de toutes leurs décisions !

@iMart1n @AppStore 😍😍😍

🇺🇸 @AppStore App of the Day, let’s f****** go 🚀🌕. Proud to work with such a talented team ! Come join the party amo.co/jobs/ #OnTheAppStore

Next Thursday I'll be speaking about how we improved the way we run @Amplitude_HQ experiments at @photoroom_app ! luma.com/ffua5h2p

Let’s go @photoroom_app 📸

We're now vibe designing 3d objects at @photoroom_app ! I love this

👉 Experimentation frameworks embed assumptions. 👉 Those assumptions don’t always match your product strategy. 👉 And it’s worth pushing until what you measure reflects what you’re actually trying to improve. I wrote down everything we learned here  photoroom.com/inside-photoro…

I just got a Christmas gift from @Amplitude_HQ 🎄

With this approach, each workflow initiation is treated as an observation, so experiment results reflect how often workflows succeed — not just whether a user ever succeeded once. The real lesson though isn’t the feature.

But the previous methodology made it hard to reason about. After many discussions with the Amplitude Experiments team (thanks Rohit Ramaiah and Eric Metelka), Amplitude just launched a new option that finally fits our needs.

We realized that the default way experiment funnels were calculated did not align with Photoroom’s product optimization goals. At Photoroom, we don’t optimize for one successful user. We aim to improve every step of our users’ workflows — for every workflow.

In my first two articles, I explained why Photoroom uses Amplitude sequential testing and how feature flags are the backbone of our experimentation system. In this third one, I share the calculation limitations we ran into.

@photoroom_app @BambulabGlobal @BambulabGlobal what do you advise to prevent spaghetti?

To me, the best perk we have at @photoroom_app is a @BambulabGlobal 3d printer anyone can use. It's like Christmas everyday

@rdethiersant @WorkwithIsland Bravo!! ☺️

5 ans pour confirmer notre vision. Une nouvelle marque pour l'affirmer. En 2020, au lancement de notre précédente identité, nous étions 7 ou 8 au 122 rue d'Aboukir chez @WorkwithIsland et nous ne vendions nos cabines qu'à nos copains du Sentier.