Alexandra Johnson
@alexandraj777
Followers
2K
Following
1K
Media
75
Statuses
892
Co-founder and CEO @RubberDuckyLabs (YC W23) Prev: @southpkcommons member, @SigOpt + @gantry_ml founding teams, AI PM @Citrine_io, SWE @Polyvore, @RenttheRunway
San Francisco
Joined November 2009
Congrats to the @Bauplan_labs team on the their launch!!
π Introducing Bauplan A serverless, code-native platform for building data and AI pipelines β directly on your object store. No clusters. No notebooks. No GUI based workflows. Just Python + SQL + S3. π https://t.co/aJGGCltlAg
0
0
6
Very excited to be onboarding with @CalvinnChenn at @fetchrhq -- AI has brought so much energy and optimism to online shopping! Can't wait to see the clothing picks.
2
2
16
RTing for reach
Need feedback from Cloudflare Durable Objects/PartyKit/Agents devs: I keep getting people DMing along the lines of "@ActorCore_org is so much easier to use than Durable Objects", but I don't know how to best refine this message. How would you phrase it?
0
0
2
Announcing: π¨ Rivet Studio v1.0 A powerful development environment for @rivet_gg Actors & @CloudflareDev Durable Objects for realtime apps. π§ View/edit actor state π Test RPCs like Postman π Explore live connections π₯ Hot reload everything β studio (dot) rivet (dot) gg
2
12
78
Praying for a full recovery π This attack is terrifying
i was recently discharged from the icu at the sf hospital after traumatic brain injury due to a man hitting me on the back of my head with a metal pipe, completely unprovoked, in the mission in san francisco. this man is known to the police since heβs done this before to other
0
0
4
Hey Claude, can you generate me some images that show how you view the aesthetics of NYE outfits?
1
0
2
Startups launch regardless if its near Christmas or not.
Rivet Actors are the fastest way to build and scale realtime apps. Available today π
0
1
18
Orrickβs Laura Barr connected with @alexandraj777, the CEO and Co-Founder of Y Combinator-backed @RubberDuckyLabs, for an inspiring conversation about deploying AI to improve recommender systems. Watch the video to learn more.
1
5
5
New post on @RubberDuckyLabs: @alexandraj777 and I have heard a lot of horror stories of ML and senior leaders completely talking past each other when it comes to metrics. To optimize on CLTV, stop trying to iterate on CLTV! Use tactical goals instead.
1
1
6
What makes some RecSys teams operate so much more effectively than others? It's all about unblocking the fast iteration loop.
To paraphrase Russian author Leo Tolstoy: All happy RecSys teams are alike; all unhappy RecSys teams are unhappy in their own way. Many struggle with common pitfalls: slow iteration, poor instrumentation, and unclear success metrics. https://t.co/RokR5orQJj
1
3
6
Follow along for more! And reach out if the ideas in the blog post resonate with you π
0
0
2
That's easier said that done - at @RubberDuckyLabs we've seen RecSys teams encounter unique challenges prototyping and shipping. Our new blog post series shares real-world best practices of successful recommender systems teams.
1
0
1
A successful team doesnβt succeed by being right every time: Most ideas end up on the chopping block. What differentiates a successful team is the ability to experiment with ideas quickly to find the few that truly make a difference.
1
0
1
To paraphrase Russian author Leo Tolstoy: All happy RecSys teams are alike; all unhappy RecSys teams are unhappy in their own way. Many struggle with common pitfalls: slow iteration, poor instrumentation, and unclear success metrics. https://t.co/RokR5orQJj
rubberduckylabs.io
To paraphrase Tolstoy: All happy RecSys systems teams are alike, all unhappy RecSys teams are unhappy in their own way. What differentiates a successful team is the ability to experiment with ideas...
1
2
6
At the last Heavybit summit, @stefkrawczyk let me drag him into leading an MLOps session together. This snowballed into a longer conversation musing on the nature of MLOps over the last decade π
We at Heavybit fully expect machine learning to be the next great driver of change in operations. But MLOps seems to have fallen short of expectations. Why? Hear from @azimman, @alexandraj777, @AndrewFong, and @stefkrawczyk.
1
2
10
The incredible @Alexandraj777 CEO of @RubberDuckyLabs back at @SCSatCMU @CarnegieMellon for her 10th reunion during Spring Carnival visiting Women@SCS students in front of my old office in Gates! None of the undergrads here were born when I started Women@SCS 25 years ago!
0
2
13
Personalization and recommender systems LFG! Think through how you can personalize every surface of your product to reflect your users' preferences, from copywriting to product recommendations.
As AI models get to some baseline level of intelligence, the moat might transition to human preferences -- whether the model is giving you the answer you want, not just the most frequently preferred answer. If this happens, companies who already have personalization models on
0
0
2
This is so close to the truth it's scary -- Rubber Ducky Labs's name was inspired by the giant rubber ducky art exhibit that visited Pittsburgh in 2013, almost exactly a decade ago!
0
0
5