Engineer on Google Spanner, formerly @cloudera. Founder of @ApacheKudu. PMC @hadoop @HBase @ApacheImpala. @BrownCSDept alumnus, wannabe pianist and cellist

Joined October 2008
47 Photos and videos
Todd Lipcon retweeted
Unleash the power of your data with our new Spanner features! Introducing Spanner Graph 🤝 Full-text search 🤝 Vector search—the ultimate trio to supercharge your app development → goo.gle/3LMqcLZ
1
17
38
7,512
Todd Lipcon retweeted
📢 Introducing *NEW* database capabilities across the Google Cloud portfolio! From Spanner Graph to Bigtable SQL, find out how these innovations help you build the next generation of AI-powered apps → goo.gle/3LM6fVB
3
14
70
13,186
24 Jul 2023
Anyone else having flashbacks to early X11 experiences with this new twitter logo? looks an awful lot like `xlogo`
5
2
16
2,213
17 May 2023
uarch experts: do not-taken forward branches (eg to exceptional paths never expected to be taken) pollute BTB? Is it better to transform the code to produce an invalid pointer load and catch it in a signal handler? (cc @trav_downs)
1
2
2,318
15 May 2023
Happy to see the open source release of the Python "causalimpact" library: github.com/google/tfp-causal… - very useful tool for evaluating impact of software optimizations when A/B simultaneous tests are not possible.
1
1
22
3,511
11 Mar 2023
Any doctors on my Twitter who want to tell me if this ear is infected?
3
1,741
11 Mar 2023
photos.app.goo.gl/xQxhFXZU7W… for the full immersive (grossly hairy) experience

1,073
5 Feb 2023
My favorite Cellar Tracker feature is definitely "Un-drink Bottle". I wish that option were available in real life.
1
1
7
4,902
9 Jan 2023
Does anyone know where to go in the bay area for urgent (but not emergent) non-COVID health issues when COVID ? Wife dealing with some acute pain,been to ER twice who told her to go PCP, then just had the worst imaginable experience at @SutterHealth
4
2
6
4,975
9 Jan 2023
(happily scheduled an appointment in internal medicine despite being COVID . We showed up, called ahead to ask if we should just go on in, and they said "sure!". Checked us in. Then called us on the phone, while we're in the waiting room, to cancel on us. wtf)
2
2
1,532
9 Jan 2023
Seems ERs won't help for non-emergency issues, primary care won't see COVID ... urgent care questionable ability to do imaging or deeper diagnosis. You'd think in the bay area it might be possible to get better care than this...
4
2
1,509
8 Dec 2022
It's funny that we now have AIs that can pass Google algorithms interviews, but apparently don't yet have AIs that can take my PDF hotel receipt with 30 line items and itemize it into Concur. Pretty sure the 10 minutes I spent entering $2 tax fees separately was not ROI.
3
2
23
20 Oct 2022
Performance people: is there a secret MSR or something that can clear most/all uarch state and caches? Would love to be able to more effectively benchmark "worst case" executions of a code path
3
1
5
20 Oct 2022
1
12 Oct 2022
Any Peninsula / Bay Area folks have suggestions for pediatric urgent care that don't involve waiting for hours? No one has appointments and "wait for 3 hours with a very sick 2 year old" is not appealing.
6
12 Oct 2022
The dirty capitalist in me wonders why there isn't a place I can pay too much money and get seen ~right away. Or at least a place where the waiting line is a virtual one and I can just get in the car 30min before our slot is up.
2
1 Sep 2022
When I interviewed with the Google Spanner team, I asked: "is Spanner done? What's left?" Turns out the answer is (obviously) that infrastructure is never done, and huge changes are still possible. cloud.google.com/blog/produc… details one such transformational change
1
6
46
1 Sep 2022
This migration was already under way when I joined 2 years ago, but I feel proud to have contributed to a few fixes and improvements in my time on the team. If you're looking to make an impact on what's probably the largest OLTP system in the universe, come interview with us!
4
11
29 Jul 2022
Anyone aware of research on IO-scheduler-aware caching policies? ie an algorithm that understands that miss cost of 5 adjacent small pages is approximately the same as miss cost of a single page due to IO coalescing? Surprisingly turning up little on this topic
1
1
11