📌We want to deliver the most accurate transcriptions for our users. So, after developing it for ourselves, we ran a test...
📊 We used to compare the error rate of 's Whisper model against Google's flagship Speech-to-Text.
Follow
Statsig
@statsigio
We help people use data to build better products.
Statsig’s Tweets
. recently changed feature flag providers to , and we couldn't be happier!
c5r.medium.com/feature-flaggi
PS, if you like this sort of thing, we're hiring 😉 DMs are open!
1
4
14
Way to go, Anu! 🤩
Quote Tweet
Transitioning to our panel of What Enables Application Intelligence - we've got @peteratmsr moderating: @JustinBorgman, @anusharma & @ajratner here with us today!
@starburstdata, @SnorkelAI, @statsigio, @Microsoft #IASummit #IA40
7
I'm looking forward to a great discussion about experimentation next Week at 's Data Science Meetup next week in San Francisco. You can find the sign-up link here if you're interested in coming and networking:
3
2
19
Craig is truly living up to our signage today.
1
7
Day deux of #dbtcoalesce in New Orleans is rolling. Give us a shout if you want to meet—we’re here all week! 😊
1
6
Have you ever had a rollout disaster?
Replying to
We use for feature flagging, and while we'll lean into pure experimentation more in the future, our eng team actually leverages it for several other things:
1
2
8
The next time someone says your metrics charts are "music to their ears," be sure to press the play button. 😏 Introducing Statsig's musical charts. 🎶🐰🥚
0:30
127 views
3
10
Anyone who has ever accidentally typed "produce" instead of "product," this one's for you. 🥂
1
8
Hooray! 🎉 Today, 50 U.S. governors signed the Compact To Expand K-12 Computer Science Education, showing a united front. There is broad consensus that CS belongs in the K-12 core curriculum. Please spread the news!
27
166
445
Show this thread
Wow! The CEOs of the largest companies and nonprofits in the United States have joined forces to make computer science a basic part of the K-12 curriculum. See who’s on board:
115
807
1,478
Show this thread
Elon Musk's a Frequentist!
Twitter claims their spam account rate is less than 5%. Elon Musk is validating this by sampling 100 Twitter accounts (a Frequentist strategy!). But couldn't we get 7 fake accounts by chance? What number of fake accounts can invalidate this claim?
1
5
6
Statsig raises $43M Series B round for rapid, live testing and analysis of new product features
1
4
8
It’s a big day at ! We’re proud to announce our $43MM Series B, led by & .
This gets us closer to our mission of rapid, high-quality product experimentation for teams of all sizes.
Forbes broke the news here:
3
8
25
Happy holidays from the team at Statsig!
7
We started testing out at Brex and as someone who worked on experimentation tooling at FB it feels like home. 🤩
3
20
Running experiments and want to make sure your results are replicable and reliable? Statsig's Export Report feature allows downloads of both summary and raw data.
1
3
Another Monday, another new feature!
Organizations is an easy way to manage projects across a company, handle access controls, and get better visibility into billing + membership.
4
New feature alert! Statsig now supports time series- view your feature results for specific days, or select one metric for results over time.
1
6
1
5
Did you get a chance to see us at AWS Re:invent?
8
How Statsig runs 100x more cost-effectively using Amazon ElastiCache for Redis
1
2
How Statsig runs 100x more cost-effectively using Amazon ElastiCache for Redis
1
2
Introducing Statsig as a Segment Destination
3
New blog up on tips + tricks for pipeline generation-
1
Our data scientist Tim walks through Feature Gates and Pulse in our new video-
2
User-Level vs. Device-Level Experiments with Statsig by Maggie Stewart
2
New blog up- it's quality week at Statsig!
medium.com/p/dab8890008a8
1
New blog up: Embracing overlapping A/B tests, and the danger of isolating experiments.
4
New blog up: A/B Testing for Dummies
3
Thought from one of our product managers today- we think that all of our ideas are good. 1/3 of our ideas are actually good. This is why we need A/B testing!
1