Rif Kiamil ( Google Developer Exper, DBT & LookerML )
LightNews LightNews
banner
rifkiamil.bsky.social
Rif Kiamil ( Google Developer Exper, DBT & LookerML )
@rifkiamil.bsky.social
14 followers 160 following 20 posts
a Google Developer Expert & proudly sporting a trophy case of accolades from Microsoft.

"You haven't truly lived until you've panicked the NT Kernel with NMI Interrupt." Rif 2012 - TAP AX6 & WAVE13
Posts Replies Media Videos
rifkiamil.bsky.social
How do you make the Borg more efficient?🤖 #AlphaEvolve!🐝💻 Google's Borg paper (circa 2015) showed how they herded millions of servers like one big hive mind. AlphaEvolve was given the task to solve & gave Borg hive mind back 0.7% of its compute 👍 #deepmind storage.googleapis.com/deepmind-med...
May 15, 2025 at 3:00 AM
rifkiamil.bsky.social
See more on research.google/pubs/ - Read Dremel paper research.google/pubs/dremel-... or get notebooklm.google turn the paper into a podcast.
April 24, 2025 at 6:16 PM
rifkiamil.bsky.social
Unleashing graph analytics in BigQuery & Spanner Graph: Graph databases reimagined at #GoogleCloudNEXT youtube.com/playlist?lis...
April 23, 2025 at 4:15 AM
rifkiamil.bsky.social
🐼 🐻‍❄️ Allegory of Pandas & Polars - Long ago, in the age of data’s infancy, there was the 🐼 , a peaceful yet slow creature, tending to its small forests of structured information.... medium.com/p/a7c1be6f87fe #Polars #Pandas #Friesland #Rust #Python #DataFrame
🐼 🐻‍❄️ Allegory of Pandas & Polars Long ago, in the age of data’s infancy, there was the 🐼 , a peaceful yet slow creature, tending to its small forests of structured information. It thrived, for a time. But as the world grew, so did its hunger for data. The ice began to creep in. The panda, unprepared, struggled in this new world of parallelism and scale. From the cold lands of Friesland, a new beast arose, the mighty polar bear “Mei de kop yn ’e wyn”! 🐻‍❄️ faced the elements head-on, faster, stronger, honed by the ruthless efficiency of Rust. It did not forget the panda’s ways, but it sought something greater. Thus, Polars was born, an heir to the old, but a ruler of the new…
April 21, 2025 at 11:36 AM
rifkiamil.bsky.social
I was asked,"What is MS V-Order?" "V-Order sorting has a 15% impact on average write times but provides up to 50% more compression." Data is likely delayed to accumulate more before being sorted in different ways, before dictionary & run-length encoding. The sort has a significant effect on & RLE.
V-Order sorting has a 15% impact on average write times but provides up to 50% more compression. What’s likely happening is that the data is being delayed slightly, allowing more of it to accumulate before encoding. Additionally, the data is being sorted in different ways to optimize encoding techniques such as dictionary encoding and run-length encoding. These concepts are covered in Inside the VertiPaq Engine.
February 28, 2025 at 1:21 AM