Monica Rogati
@mrogati
Followers
50K
Following
9K
Media
120
Statuses
2K
Data Science & AI advisor; fractional CDO. Former VP of Data @Jawbone & @LinkedIn data scientist. Equity Partner @DCVC. CMU CS PhD.
Sunnyvale, CA
Joined November 2008
The AI pyramid of needs, or why you need a solid foundation for effective AI and Machine Learning:
42
644
1K
I've always been passionate about data quality -- that's where the best war stories are, too. My friend @jeremystan is writing book on it and you can read the first chapter now:.
šĀ HOT OFF THE PRESS š. Weāre excited to share an exclusive first look at @OReillyMediaās new book Automating Data Quality Monitoring at Scale written by @jeremystan & Paige. Chapter 1: The Data Quality Imperative is out now. DownloadšĀ #dataquality š.
1
3
17
RT @criccomini: Hi, this is your annual reminder that @squarecog and I wrote a book a year ago. Reviews are in and they're pretty solid, ifā¦.
0
9
0
RT @jeremystan: We just released our first post in a series on monitoring metrics in @anomalo_hq. Visualizations are š to doing this well:ā¦.
0
8
0
If these SOTA prompts remind you of SEO, thatās because theyāre both iteratively manipulating inputs to a black box in an attempt to steer it.
@chrisstanchak @karpathy @MichaelTrazzi Hereās an example of a āstate of the art promptā, in my opinion, it involves a lot of overloading, and sometimes adding in things you donāt exactly want like ādeep sea creatureā (or whatever else) to find more interesting forms and textures . Iād love to see other approaches
1
0
11
RT @josh_wills: I'm very excited to announce that I will be teaching a.course on Data Engineering for Machine Learning in September! (1/5)ā¦.
0
35
0
Bindu, can you please explain why you took my work, replaced my watermark with your logo, and tweeted it as your own? Many people I respect follow you so Iām trying to think of a good explanation but coming up short.
1
6
31
RT @mejoeyg: .@joe_hellerstein, @vsreekanti, @cgwu0530 and I have been working on a new project (and company) to simplify infrastructure foā¦.
0
20
0
RT @jeremystan: We don't store data in warehouses. We operate data factories without quality controls. Then we fight fires when dashboardsā¦.
0
11
0
Take a look if you're working with Machine Learning for NLP. Fun challenge and a change of pace if you're advising, consulting or taking a break. Slides and lectures already available as a starting point!.
We are on the search for a lecturer to teach NLP 243, Machine Learning for Fall 2022!. If interested, send an email to nlp@ucsc.edu indicating your interest and qualifications, with subject line "Teaching NLP243". Syllabus:.
0
0
2
RT @criccomini: š The project I've been working on is now open source!. Open Robo-Advisor is a Python library that acts as an advisor š¤ forā¦.
0
20
0
RT @myelbows: Beyond excited about ā¦@streamlitā© + ā¦@SnowflakeDBā© . Itās a win-win-win for the community, for our technology, and for customā¦.
0
20
0
You might think only applies to big companies with complex planning and metrics, but I'm surprised how often people don't pause to wonder 'OK, what if this works?' before getting started.
0
1
5
A question I ask to prioritize data projects: . 'Imagine you're done. It took 4 months. It works OK. Now, what metric has improved? By how much? Was it worth the effort & opportunity cost?' . More on assessing impact from @jikechong & @yuec's book on leading in data science:
6
11
57
Same. @TheBeastAcademy is one of those rare products that immediately turn you into an evangelist. After trying so many math apps that were based on rote memorization or games w/ the occasional addition question, BA was the only one I felt could inspire a love of math.
Our boys are doing AoPS (or Beast Academy) for their math curriculum. 1 year in, they are both loving it, are developing strong mathematical thinking and interest in exploring more. I attribute all this to AoPS.
0
4
22
My favorite data science algorithm (division) saves the day ā this time as the final step of automated bad data root-cause analysis by @anomalo_hq. Intuitive ā”ļø easy to trust, which is key for this application.
Row level insights are key to explaining data issues:. - expected or not?.- cause for concern?.- where did it happen?.- why?. Our root cause analysis does this automatically, every time. New post by machine learning engineer @VickyAndonova explains how.
0
3
8
For most ML systems, data quality is *the* impact bottleneck. It sneaks up on you while the data team is busy modeling; when discovered it's dismissed as a 'bug, fixed' vs. a systemic issue. Impressed with what @jeremystan and @eshmu built so far & their vision; congrats!.
Big news from @anomalo_hq today. Huge thanks to the team, our investors, and our amazing customers for making this possible.
0
4
15
What feature stores are people using these days? Are you hĢ¶aĢ¶pĢ¶pĢ¶yĢ¶ less annoyed with your current solution vs. others you've tried?.
2
2
7