Conference that could have been an offsite.
Offsite that could have been a workshop.
Workshop that could have been a meeting.
Meeting that could have been an email.
Email that could have been a slack message.
The Slack messageโฆ
@pritul.bsky.social
16 years Data scientist ๐จ๐ปโ๐ป Data science career coach ๐ค A/B testing and Causal inference ๐งช Big-tech Experimentation Infra ๐ CommLead @ "Experimentation & Causal AI" Skool ๐ Create and Curate Data Memes ๐ผ๏ธ
Conference that could have been an offsite.
Offsite that could have been a workshop.
Workshop that could have been a meeting.
Meeting that could have been an email.
Email that could have been a slack message.
The Slack messageโฆ
I am pretty sure that the best of the time and project management junkies would not be able to optimize their skills with a toddler at the Disney's Magic Kingdom park.
04.04.2025 13:55 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0Grandparents after OpenAI launching OPERATOR
24.01.2025 15:42 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0๐ It is Friday the 13th and I had a nightmare!
I turned into a ....
Bayesian ๐
I kept saying '95% probability of treatment winning' and stopping tests whenever I felt like it.
WTF... What a monster! ๐น
These are just jokes. Just like 'uninformative priors' ๐
#databs
Thug life ๐ of a data scientist.
Sample size calculator: "2 weeks"
Stakeholder: "I need it in 3 days"
Data scientist: "Make it 4 weeks and I'll let you peek once"
#databs
๐ thatโs an eye opener.
21.11.2024 16:19 โ ๐ 1 ๐ 0 ๐ฌ 1 ๐ 0I turned 40 yesterday. This morning I woke up with immense bout of wisdom that all life's mysteries seem clear... except why I need to type GROUP BY age, region, product when SELECT already knows what I'm aggregating. ๐ค
#databs
How to p-hack like a pro!
#databs
Added to the ๐ฆ CRO Starter Pack this weekend:
@tlckelly.bsky.social, @simoahava.bsky.social, @vreeman.com, @els-aerts.bsky.social, @smarslan.bsky.social, @clarquey.bsky.social, @chrisgibbins.bsky.social, @timthijsse.bsky.social, @hellemans.bsky.social, @julianajaxx.bsky.social, @pritul.bsky.social ๐ช
Data analyst provides descriptive understanding of the business. Data scientists take that knowledge and build statistically sound models for different areas of the business - churn models, predictive models, causal models, etc. These directly inform new data products and business decisions.
17.11.2024 19:46 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0Early signs your 4 year old will become a Data scientist.
Me: How many toys do you have?
4yo: Not enough
#databs
โEveryone has a SQL plan until their query optimizer punches them in the face.โ
Michael Tyson, Senior Data Engineer.
#dataBS
Slight modification.
-learn ML and become Data Scientist (no need to deploy to prod)
-learn ML+infra+engineering and become MLE (deploy scalable algos to prod)
Everyone's posting about Tyson vs Paul. I didnโt watch having zero interest in boxing. Yet feeling compelled to post something about it. In behavioral psychology, this is called โconformity biasโ. Same reason stakeholders suddenly want features their competitors just launched. #databs
16.11.2024 21:51 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0For two days in a row I have battled with vercelโs AI app builder. That thing is not smart to even build a simple react app. The app works only in its env. Pushing to prod breaks it.
16.11.2024 21:00 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0At 25 discovered Twitter. At 40, Bluesky! New feed, new algorithm overlords to appease๐
16.11.2024 20:55 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0A two-panel Drake "Nah/Yeah" style meme featuring Star Trek character Geordi La Forge. In the top panel, Geordi is holding his hand up in rejection to "yelling at models (fashion)". In the bottom panel, he's pointing approvingly at "yelling at models (AI)". The meme plays on the double meaning of "models" in fashion versus artificial intelligence contexts.โโโโโโโโโโโโโโโโ
Back propagation Out โ
Back-hand propagation In โ
#databs #meme
May I request to be added to CRO starter pack please?
16.11.2024 10:55 โ ๐ 1 ๐ 0 ๐ฌ 1 ๐ 0Fell for insights, stayed for Stockholm syndrome.
Happy Friday!
#databs
Can you add me as well?
14.11.2024 03:36 โ ๐ 0 ๐ 0 ๐ฌ 1 ๐ 0Thats a cool functionality! Thank you.
14.11.2024 03:33 โ ๐ 2 ๐ 0 ๐ฌ 0 ๐ 0Who should I be following here? #databs
13.11.2024 22:33 โ ๐ 2 ๐ 0 ๐ฌ 1 ๐ 0