Activity
1K followers
-
Wei Xiao shared thisSomething my team is working on ...Wei Xiao shared thisIncreasing performance and reducing the cost of MXNet inference using Amazon SageMaker Neo and Amazon Elastic Inference #aws #sagemaker #machinelearning
-
Wei Xiao shared thisWei Xiao shared thisAWS launches initiative to accelerate COVID-19 diagnostics, research, and testing
-
Wei Xiao liked thisWei Xiao liked thisA post recently written by my friend and former colleague, Mohamed Fathalla, has prompted me to collect my thoughts about why the expectation that AI will revolutionize Software Engineering is going to result in disappointment. As the content turned out to be too long for a post, it became my first article:
-
Wei Xiao liked thisWei Xiao liked thisHere we go again! I resigned my position as Principal Engineer at AWS with a last day of May 23rd 2025. This was a much shorter stint than I anticipated. I have to assume I was not ready to resume my career and maybe I should have enjoyed a longer break last year. AWS did nothing wrong, this is all on me and I can’t shake the feeling I squandered a very good opportunity, but it is what it is. Anyhow I spent 6mo or so in Bedrock/Sagemaker inference hosting and I grew convinced I simply failed to lead. You see, being a leader at AWS is not just something associated with the title and seniority, you need to embrace your team mission and goals and drive the business to success. I know I was useful to the team, many team members told me so explicitly before I left. I even had peers question why I was so hard on myself when I was telling them how nothing I did felt good enough to me. But all along I still felt I was no leader. Before leaving AWS I decided to try a lateral move first, so I changed team and domain but within mere weeks I started to feel the urge to spend time with family and so I resigned. The new role had me back more into direct coding, given a build could take 30mins or longer I’d leave the office to pace myself outside. During these side walks I started to feel that maybe I should spend more time with family especially with my kid before he wouldn’t want dad around anymore. My daily routine during this last AWS stint would consist of waking up at 5:30-ish for some “me” time (read news, check investments, read something fun) before hitting the gym at 6:30/6:45, back by 7:45 and in office by 8:30. Work until 6:30/7:00PM, come home, have dinner and crash on the sofa by 7:30PM pretty tired … needless to say no time to play with my kid. Weekends were better but barely as I felt bad for what I perceived I failed to achieve at work. Maybe my routine is all wrong, foregoing quality time with family during the week assuming I would catch up on weekends while instead I was sulking, in hindsight seems a very bad idea. Those strolls between builds put this into perspective, I obsessed over it for a week then the answer was obvious. I felt ashamed I left the new team so soon, the new manager and the few colleagues I interacted with all understood and were extremely gracious with me. The plan now is to enjoy summer and be just dad. This first week of being home already brought positive change. I can afford to go to the gym later at around 8, I do get to sleep longer. This enabled me not to crash right after dinner and because I am home even before dinner my son is asking again that I play with him …. nothing glamorous or mind blowing but it feels good to be able to be dad. It’s indeed the little things. We will see what the future brings, if this is retirement or just another break I do not know, but for sure if I get back into the workforce I better learn how to balance work, life and health, thus far I haven’t done a good job.
-
Wei Xiao liked thisWei Xiao liked thisI had a great time at Amazon between 2010 and 2015. After almost 10 years later, I was honored with an offer to join the mothership again. What has changed in 10 years? … Me :) It’s been almost a decade constantly reflecting and using Amazon leadership skills in multiple situations over my career - most of the time in silence, so it did not trigger any allergies. This time I am focusing so much more on leadership skills than in my first run. Working at Amazon is like getting a multi-year MBA that is not only free but you get paid for. It is a big responsibility to be back. Tomorrow I will have another Day 1.
-
Wei Xiao liked thisWei Xiao liked this“All good things must come to an end”… Nine years ago today I was walking around San Mateo interviewing with Mike Speiser praying that I didn’t screw it up. Snowflake sounded cool but honestly I just really really really wanted to work with Chris Degnan again. Mike of course got me totally jacked up “Hadoop is dead!”. When we got back, I had lunch with Benoit Dageville, Thierry Cruanes and Marcin Zukowski. It was a little intimidating meeting with three of the smartest humans on earth but they were so kind, smart, driven and passionate. The day was sealed with a huge smile from Nancy Venezia. Leaving, I thought “Wait…I get to work with Chris AND the other people are awesome too?! Dream Job!” I was so fired up... Snowflake redefined the definition of dream job for me. Obviously, Snowflake turning into a generational tech company is pretty rad, but the thing I’ll miss the most is the people. I got to work with people who were focused on the mission, smart, insanely hard working, and unique. I got to work with my friends…that’s the dream. So what’s next for me? 1. Spend time with my wife and two little kids 2. Focus on health (startups are worth it but not healthy) 3. I’m not fully retiring, I’m just going to transition into a new type of work that better fits priority 1 and 2. I’ll talk more about it after the holidays but think startups and this massive new application ecosystem Snowflake has created. Thank You - - Thank you Chris Degnan. I have so much respect for what you have accomplished. You are the best sales leader in IT and an even better person. - Thanks you to the geniuses/artists past and present who built, designed and supported one of the all time greatest products in IT. I was lucky to sell a product that made customers happy. You are the BEST engineering and product team ever assembled. Benoit Dageville Thierry Cruanes Marcin Zukowski Grzegorz Czajkowski Leo Giakoumakis Ashish Motivala Abdul Q Munir Allison Lee Christian Kleinerman Artin Avanes Jonathan Claybaugh - Thank you Brian Kelly for supporting me and my vision for the Data Cloud team. You are so talented, Snowflake is lucky to have you. - Thank you to the entire Snowflake Executive team past and present. Not once in nine years did I ever worry that the ship wasn’t sailing in the right direction. Bob Muglia Denise Persson Frank Slootman Michael Scarpelli ***Last but not least…In fact the reality is that I owe 100% of my success to you…THANK YOU to every person that took the chance to work for me. The West sales team was/is the best. My enablement crew was so good. The Data Cloud team was legit the all star team. I owe each and every one of you a huge debt, don’t hesitate to let me know how I can repay you. I’m signing off with a picture of one of my happiest times at Snowflake. - Kyle
-
Wei Xiao liked thisWei Xiao liked thisKanari Digital Inc., began its journey about 2 years ago with a mission to help all businesses respond to business events rapidly and in an automated fashion. Our objective is to have a large impact in the digital transformation of businesses leveraging near real-time data and automation. We bootstrapped, remained lean and customer-focused and most importantly put learning and having fun at the core. Today, we have made a decision to join UiPath to help realize their vision of the fully automated enterprise. It is the best way for us to realize our ambitions. Our team is becoming part of UiPath to help leverage data and take automation to every part of an enterprise. This was an easy decision for us as UiPath is the recognized leader in enterprise automation software. We are looking forward to accelerating business transformation for UiPath customers across the globe. Many people helped us along the way and pushed us to do better. You know who you are. We thank you all for lending us your ears and giving us the most precious thing you have – your time.
-
Wei Xiao liked thisWei Xiao liked this*dj air horn* Finally published! My magnum opus about building multi-tenant services at Amazon: https://lnkd.in/gS2yGyp . This started out as a simple article about API rate-limiting, but the really interesting bits ended up being about how we think about multi-tenancy. #cloudcomputing #awscloud
Other similar profiles
Explore more posts
-
Josue “Josh�� Bogran
In today’s news: -Microsoft announces Clippy Copilot, promising investors to be the one Copilot product that customers actually use. -In efforts to spread out regional demand, AWS announced it will rename all regions to “AWS US East 01”. -Meanwhile, GCP dropped a marketing campaign to remind the market that it is a viable 3rd option. #satire
-
Robert Hundt
We are reimagining PyTorch on TPU (https://lnkd.in/gCrM-TZD). This approach still calls XLA but in a profoundly different way, enabling a solid eager mode. In the limit, this new approach will enable a true JIT approach with recompilation and re-optimization, online feedback-directed optimizations, avoidance of OOMs, support for HW fungibility, hyper-optimizations of kernels, and more - all with close to zero code changes!
2 Comments -
Paul Dudley
Streaming 400,000 tables into Snowflake with Streamkap 🤯 When I first met Casey Lavallee at Limble he shared the challenges with their current architecture: -400,000+ source tables from MariaDB replicated to Snowflake -A DIY pipeline that was expensive and unreliable, requiring dozens of hours of maintenance per month Casey had looked at other commercial ETL and streaming options, and none were able to meet the challenge. The Streamkap team was able to get Limble migrated quickly and enabled: ✅ Table Collapsing: Merging 400k+ tables into manageable schemas before they hit Snowflake. ✅ BYOC: Kept everything secure in their own AWS environment. ✅ Snowpipe Streaming: Ditched the expensive warehouse loads. The Impact: 📉 Ingestion Costs: Dropped by ~40x. ⏳ Maintenance: From 24 hours/month down to 2. 🚀 Latency: Real-time data availability for product analytics. Give Streamkap a look if you're solving for large scale multi-tenant or single tenant architectures. Check the comments for the full case study.
8 Comments -
Sanchit Narula
The best advice I got on my promotions as a software engineer was from a Principal Architect at Amazon after he told me his story of getting promoted to Principal, and this has stayed with me my whole career. He said, very simply: > Your architecture can be a love letter to your technical brilliance, > but if it does not move customer and company goals, it will not move your level. That hurt a little. Because I recognised myself in it. A few years ago I wanted to design something with every shiny thing in the book: API Gateway, Lambda, DynamoDB, event bus, the works. On paper it was beautiful. In reality it was expensive, slow to ship, and did not match the skills of the team that would maintain it. So we shipped a hybrid architecture instead. Fewer moving parts, more reuse of existing services, cheaper to run, easier to hand over. Over time, here is how my thinking about promotions changed. 1. Start from the business goal, not the tech stack Before talking about Kafka vs SQS or MySQL vs DynamoDB, answer this first: What money, risk, or customer pain will this remove? Promotions track that, not the number of services in your diagram. 2. Ship finished products, not clever demos Leaders remember the feature that went live, stayed stable, and earned or saved real money. Drive the entire lifecycle yourself: design, build, launch, oncall, cleanup, docs. 3. Make the team faster, not just yourself Tools, scripts, templates, better CI, better dashboards. If people quietly say “things move faster when you are on this” your manager already has half a promotion case. 4. Do your share of the unsexy work Oncall, migrations, bugs, incident reviews. Senior folks who avoid this slow the entire org. Senior folks who lean into it become trusted very quickly. 5. Grow people, not just systems Mentor juniors, run design reviews, write clear docs, set standards. When your manager sees that you are already acting like the next level, formal promotion becomes a catch up, not a favour. 6. Be proactive about your promotion, not obsessed with it Have periodic one to ones. Talk about where you want to go and ask what evidence your manager would need to support that. Remember your manager has ten other people to manage. It is your responsibility to make your impact visible. Do not be desperate. Understand the timelines and work within them. As you get close to the boundary, collect the missing data points: impact metrics, incident write ups, design docs, peer feedback. Make the story easy to tell. Most importantly, work for the business, not for promotion-driven development. If your roadmap only exists to tick boxes on a promotion rubric, people can feel it. If your roadmap clearly creates value, saves cost, or reduces risk, promotion becomes the natural side effect. So instead of asking “What new tech can I use this year?” Ask, “Where can I create obvious business impact and leave the system and the people around me better than I found them?”
15 Comments
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content