Yuanqing Ye

I first joined Amazon in December of 2005. Even though I left for two years, my "Amaversary" still coincides with the end of the year and a natural moment for reflection. As I reached my 17 year mark this December, I also had the privilege of working with Amazon Careers to share some of my journey and thoughts about individual contributor technical leader and people manager career paths. In my ruminating, I put together far more content than could be used for a single article. So, I've decided to post here as a series of Q&As - in case others might find my observations, experiences, and perspectives interesting or useful. Below is the first of 6. 1. 𝘽𝙧𝙞𝙚𝙛𝙡𝙮 𝙨𝙝𝙖𝙧𝙚 𝙮𝙤𝙪𝙧 𝙘𝙖𝙧𝙚𝙚𝙧 𝙟𝙤𝙪𝙧𝙣𝙚𝙮; 𝙬𝙝𝙖𝙩 𝙢𝙤𝙩𝙞𝙫𝙖𝙩𝙚𝙙 𝙮𝙤𝙪 𝙩𝙤 𝙩𝙖𝙠𝙚 𝙤𝙣 𝙙𝙞𝙛𝙛𝙚𝙧𝙞𝙣𝙜 𝙘𝙖𝙧𝙚𝙚𝙧 𝙥𝙖𝙩𝙝𝙨, 𝙄𝘾 𝙫𝙨 𝙢𝙖𝙣𝙖𝙜𝙚𝙧? I started my career as a software engineer 25 years ago. I've worked for startups, enterprise companies, and 7 different organizations over 17 years and two tours at Amazon. I've written firmware for robotic camera mounts; built highly scalable distributed systems; designed restaurant marketing, ordering, logistics, and operations software; created automated test harnesses for mobile games; tested HD DVD encoding; helped launch grocery delivery to customers around the globe - just to name a few of the many cool things I've had the privilege to work on. I've been an entry-level software engineer, a technical lead of teams, a software manager on a couple of occasions, chief software architect reporting to a CTO, and (for the last few years) Amazon Principal Engineer directly supporting a VP with about 200 engineers within my remit. I first became a manager because I mistakenly thought it was the only way to continue growing as a leader, and in many ways I failed. I learned some very humbling lessons about leadership through that experience, which I used to be a better individual contributor technical leader on my path to Senior Engineer. My second time as manager was because my organization had a need that I could fill. I had grown enough as a leader that this time I was much more successful. I said yes to an opportunity that started with managing a small team through delivery of a complex and critical project. Once I grew a high performing team that delivered that initial project, I took on more scope and expanded into multiple teams. After a couple of years of success growing and investing in my teams so that we could deliver greater value to our customers and business, I decided to prioritize being close to my family. For me, at that time, it meant giving up my people management role and returning to the role of an individual contributor technical leader that supports a globally distributed organization. Even though I really enjoyed my time as a people manager, there is no better job for me than that of a Principal Engineer. #amazon #careerdevelopment #technicalleadership #peoplemanagement

91

3 Comments

It was interesting to study Lex from AWS together with NextWork. They came up with cool lessons that are fun to go through. Now I have finished the AI/ML block which consists of 5 lessons. Here's I used Amazon Lex and AWS Lambda to create a smart banking chatbot, BankerBot! 🤖 Here's what I did: 🚞 Created multiples slots, so my bot can handle money transfers between accounts. 🎨 Tried Lex's editor features like conversation flow and visual editor. ☁️ Used CloudFormation to automate the deployment of my bot. Learning AWS, Amazon Lex, or cloud computing? Let’s connect and explore these amazing technologies together! A big shout-out to @NextWork for their project guide. 👉 Join me at link.nextwork.org/linkedin. #AWS #AWSCloud #AmazonLex #Lambda #ChatbotDevelopment #LearningJourney

7

Take AI inferencing to the next level. Discover how to deploy and scale customized NVIDIA NIM pods on Amazon EKS. Learn how to leverage Prometheus metrics, and auto-scale with CAS, HPA, Karpenter, and KEDA. Details here

4

Our SIGMOD '24 paper explores MemoryDB's architecture, which delivers in-memory performance and strong consistency at scale. A key innovation is the decoupling of durability and replication from the in-memory database engine, which allows scaling compute and storage independently based on customer workload needs. Customer data integrity, security, and availability are paramount for MemoryDB. The paper dives deep into innovations that the team built to make that possible (a personal favorite is off-box snapshotting which allows us to create consistent database snapshots on separate clusters to avoid impacting live customer traffic and preventing resource contention). We also share hardened operational experiences from running a distributed database at cloud scale. #aws #memorydb #inmemory #databases #durability #sigmod

91

1 Comment

Last week, Steven Astorino challenged Tanmay Bakshi and I to demonstrate building and deploying an application from scratch at the IBM Canada Townhall. We made this possible by leveraging watsonx Code Assistant to accelerate our workflow, allowing us to focus on the inventive components of software development. Since the Grammy Awards are coming up, we decided to build a polling platform where the audience votes for nominees in each of the major categories so we could determine IBM Canada's predictions for the outcomes. watsonx Code Assistant helped us write boilerplate code, SQL queries, data structure transformation logic, and more! As a result, we had a fully deployed application in ~15 minutes and collected over 1400 votes from attendees. Turns out, IBMers share my opinion that Taylor Swift will win Album of the Year 😊

187

8 Comments

I’m proud to announce that after months of hard work, my team at Amazon Web Services (AWS) has successfully launched Amazon ECS Graviton-based Spot Compute for Fargate. It has been an honor to be entrusted with bringing this cost-effective solution to life. Amazon ECS Graviton-based Spot Compute for Fargate leverages the power of AWS Graviton processors, delivering superior price-performance compared to x86-based instances. Moreover, by consuming less energy, Graviton processors contribute to reducing the carbon footprint of data centers. This offering allows customers to run containerized workloads on Arm-based compute resources, unlocking cost savings while benefiting from the flexibility and scalability of AWS Fargate. To learn more: https://lnkd.in/gcvgB-GF #AmazonECS #FargateSpot #Graviton #ArmCompute #CloudComputing #Containerization #AWS #Innovation #SustainableComputing

71

2 Comments

8 years ago, I started at Google, drawn to the job and mission of supporting customers, building cloud architecture, and creating tools. Back then, our team of 35 focused on "Cloud Platform products like App Engine and Compute Engine". Funny enough, there was no mention of Kubernetes or GKE, which is now a cornerstone of cloud architecture and my specialization! My Googleversary always brings reflection, especially after the intensity of the last few years. Each year feels like a checkpoint; an opportunity to analyze what went well so far and how to build on it. I thought it would be really cool to have someone summarize for me the last 8 years working for Google, my team, and shed some light on what does a Solutions Engineer exactly do? I turned to Google's new notebookLM for this. notebookLM is a web app similar to Google Docs, but it ingests documents, sources, and videos, allowing you to ask questions in natural language. I fed notebookLM, my resume, recognitions, awards,  assessments (including growth feedback), conference notes, presentations, product requirement docs, trip reports, ~8 years worth of performance evaluation and a few rants ;-). notebookLM has a podcast feature that creates conversations from your uploaded documents. You can give it guidelines, but you can't heavily influence the output. I asked it to stop using internal code names, customers, partners, and other names. I also had to cut sections where it tried too hard, "the biggest company in X sector in Y region", that's just too revealing. The summaries were surprisingly insightful. Sometimes, it mentioned things I'd forgotten, which I then fact-checked. notebookLM responds with the exact section in the document where it found something. I appreciated this perspective on my contributions, especially the balance it showed between my technical skills and leadership, training and developing my team. This gave me valuable insights into my own strengths and how I contribute to my team. I must also say a lot of these achievements can't happen without the fantastic teammates, partner teams, customers etc that help us build the cloud community. Incredibly blessed to have spent the last 8 years doing this, and very thankful for all of it. In the end notebookLM gave me a beautiful product, but it's also not lost on me the times i've stumbled, wasn't too sure if I was focusing on the things that would ultimately make my team successful. There was a broadcast/podcast that was talking about a sport team and saying "having pressure is a privilege"...  and in the moment, I myself may not fully believe that, but it's a sentiment I'd love to embody someday! Very thankful for the 8 years of growth. *If you're an AI practitioner, and you have an idea on how to take this audio and generate something cool with it! shoot me ideas! Here's the podcast summary it generated! https://lnkd.in/gfH4zwQQ

155

7 Comments

Exciting news! I'm thrilled to announce that I'll be speaking at PG Day Chicago (https://lnkd.in/g_-vytmF) on the topic of "Probing PostgreSQL and Linux kernel interactions." My talk covers some upcoming aspects of PG17 but mainly caters to beginners dealing with PostgreSQL and Linux Kernel internals. Don't miss out on this opportunity to expand your knowledge! Check out the abstract for my talk here: https://lnkd.in/gJVcGb4u #PGDayChicago #LinuxKernel, #PostgreSQL, #PGInternals, #pgengine.

109

5 Comments

Reminder: I'm doing a webinar today at 5pm PST on creating and succeeding in high-performance teams. I'll share the tips and tricks I've learned to make my team in the top-25th percentile of all Amazon teams across key dimensions like Manager Satisfaction, Job Satisfaction, Team Culture, and others, as measured by our internal polling, while still delivering large-scale projects within 5% of target dates and driving key value for the IMDb organization. Register here: https://lnkd.in/ezEQiwsJ #TeamCulture #Career #Leadership

6

Recently, I came across an article about Amazon Prime shifting from microservices to a monolith. While it amazed me at first glance, it has also reportedly saved 90% in architecture costs. This struck a chord with me because it echoes the challenges I’ve faced in my own journey with microservices. ✅ Advantages I’ve seen with microservices: - Scalability - Flexibility - Fault isolation ❌ Challenges I’ve faced: - Increased complexity - Operational overhead - Data consistency issues There were moments when I questioned the path I was taking, following microservice architecture because it was widely regarded as a "best practice". And yes, statistics backs up this statement. This Amazon Prime example reminds me of an important lesson. There is no one-size-fits-all solution. Architectural choices should always align with the problem at hand, not just because they are old or considered the best. If you want to know more about what happened at Amazon Prime, go ahead and check out this article : https://lnkd.in/gtPgZf4J

12

1 Comment

In this episode, we discuss Learning Task Decomposition to Assist Humans in Competitive Programming by Jiaxin Wen, Ruiqi Zhong, Pei Ke, Zhihong Shao, Hongning Wang, Minlie Huang. The paper presents a method to enhance human understanding and repair of language model (LM)-generated solutions by automatically breaking down complex solutions into simpler subtasks. They introduce a novel objective called assistive value (AssistV) to measure how easily humans can repair these subtasks and validate their method through a dataset of human repair experiences. The approach significantly improves the problem-solving ability and speed of non-experts in competitive programming, allowing them to solve more problems and match the performance of unassisted experts.

2

Check out the latest in our "Debuggability" blog series! This blog is authored by Kevin Downey Jason Johl and me and talks about how we built a Debuggability Paved Road at Intuit using Ephemeral Containers and Argo Workflows. We'd love to hear your thoughts on the topic. Also, what tools do you build as-a-platform to help debug service, compute and network issues in Kubernetes at scale? #intuit #kubernetes #debug

23

In this Cloud Field Day article, Sulagna Saha discusses how Google Cloud is optimizing compute infrastructure with #Titanium and new #VM families for general and specialized workloads. #CFD20

7

2 Comments

Compute is all you need. Language model pre-training costs have skyrocketed, with researchers using scaling laws to optimize compute allocation between model size and training data, known as Compute-Optimal or Chinchilla Optimal approaches. The key insight of the Chinchilla Optimal is that many language models are being trained sub-optimally. The models shown in the referenced image collectively cost about $50 million to train at current rates. A new scaling law proposes that for transformer-based models, performance primarily correlates with total compute used, regardless of how it's distributed between model and dataset size.

21

1 Comment