f********3
发帖数: 30 | 1 在线等。
菜鸟一枚，对市场不是很了解。大家能分析下这几个组哪个更有前途吗？
之前听说AWS很不错，但是看了下技术就只列了：back end, database, cloud,
algorithms
其他几个有machine learning, data mining 啥的，听上去好像比较高大上，比较有前
途的样子？
Amazon Web Services - Infrastructure
Cloud Drive
eCFT - Computer Protection Services: 介绍说是做 large-scale distributed
security systems
Search and Discovery: 据说这个组很轻松？ |
f********3
发帖数: 30 | |
g********t
发帖数: 212 | 3 Depends on what you want to do next:
1) Stepping stone to smaller startups --> "AWS" on your resume alone will be
helpful enough. Use your spare time and your free AWS account to get
familiarized with AWS (not your daily works working on a small part of it).
Also take advantage of AWS internal metrics/on-calls to know what's
happening with the startups, which ones are good.
2) Stepping stone to other big companies such as FLG or even M$ --> try join
AWS's "front-line" products, i.e. those newly developing ones. Kinesis is
pretty good. EMR probably is still okay. There might be other unpublished
ones but if you can join you can use that to step to MS Azure or Google
Cloud
3) Want to build your own solid distributed systems experience --> DynamoDB,
S3, EMR, Data Pipeline, Kinesis all good. The real kernel part of those are
very hardcore though. But frankly I speculate if you can get into those,
most likely you will work on some kind of wrappers/peripheral components
around those cores.
4) I never heard about any AWS product called "Search and Discovery".
Probably I was wrong. Maybe you were mentioning "CloudSearch"? That is very
good too. Build search system teaches you a lot pattern re-usable. |
f********3
发帖数: 30 | 4 Great Information! Thanks a lot!
Search and Discovery 不是在AWS里面的。我表达有所不清楚，它主要是给customer提
供推荐的，就是打开amazon会把你最近看的东西放在首页或者是把你可能会买的东西放
在首页之类的。
be
.
join
【在 g********t 的大作中提到】
: Depends on what you want to do next:
: 1) Stepping stone to smaller startups --> "AWS" on your resume alone will be
: helpful enough. Use your spare time and your free AWS account to get
: familiarized with AWS (not your daily works working on a small part of it).
: Also take advantage of AWS internal metrics/on-calls to know what's
: happening with the startups, which ones are good.
: 2) Stepping stone to other big companies such as FLG or even M$ --> try join
: AWS's "front-line" products, i.e. those newly developing ones. Kinesis is
: pretty good. EMR probably is still okay. There might be other unpublished
: ones but if you can join you can use that to step to MS Azure or Google
|
p*****9
发帖数: 20 | 5 你说的这是retail的personalization&recommendation吧？最近刚拿到offer，正在选
组，感觉AWS的oncall略重啊。。
【在 f********3 的大作中提到】
: Great Information! Thanks a lot!
: Search and Discovery 不是在AWS里面的。我表达有所不清楚，它主要是给customer提
: 供推荐的，就是打开amazon会把你最近看的东西放在首页或者是把你可能会买的东西放
: 在首页之类的。
:
: be
: .
: join
|
g********t
发帖数: 212 | 6 这个问题和上面Search and Discovery的问题基本一样，所以统一说个人感觉的话:
Recommender System感觉是电商的掌上明珠之一。
以后要创业的话，个人直觉国内有很多类似的事情可以做。比如婚配之类的。
不过问题是已经做得很成熟了。Machine Learning后期的问题，基本就是调参数，很多
大量的实验，带来很小一点点改进。你没有很强的ML的背景，基本轮不到你做这些。估
计就是做这个系统上一些小修补。
下一步的话，在美国我觉得如果没有ML PHD，我觉得这个经验还是不太可能帮助去美国
这边的小公司做他们的Recommender System。比如Wish.com,Tradesy.com, etc. 他们
都有专人做了。如果回国认识一些朋友，估计可以自己做类似的。
不过就作为一段分布式系统，高吞吐系统的通用经历，后面可以有很多方向可以走。
【在 p*****9 的大作中提到】
: 你说的这是retail的personalization&recommendation吧？最近刚拿到offer，正在选
: 组，感觉AWS的oncall略重啊。。
|
f********3
发帖数: 30 | 7 但是AWS最有用好跳槽啊，苦点还是值得的！
【在 p*****9 的大作中提到】
: 你说的这是retail的personalization&recommendation吧？最近刚拿到offer，正在选
: 组，感觉AWS的oncall略重啊。。
|
f********3
发帖数: 30 | 8 哦对了，我听朋友说retail的personalization...还是比较轻松的
你打算接offer了吗？听recruiter说first come first serve,再迟有些组肯能就满了
，好纠结
【在 p*****9 的大作中提到】
: 你说的这是retail的personalization&recommendation吧？最近刚拿到offer，正在选
: 组，感觉AWS的oncall略重啊。。
|
s*********8
发帖数: 7 | 9 retail的personalization 还是有很多东西可以做的。 数据大了后做的东西就更有奥
妙了。AWS都可以用到啊。
自己体验一下recommendation & personalization, 如果你觉得还有很多不满意的地方
，那为何不试一下recommendation & personalization组呢？
【在 f********3 的大作中提到】
: 哦对了，我听朋友说retail的personalization...还是比较轻松的
: 你打算接offer了吗？听recruiter说first come first serve,再迟有些组肯能就满了
: ，好纠结
|
w*****5
发帖数: 75 | 10 Mark. 太赞了
be
.
join
【在 g********t 的大作中提到】
: Depends on what you want to do next:
: 1) Stepping stone to smaller startups --> "AWS" on your resume alone will be
: helpful enough. Use your spare time and your free AWS account to get
: familiarized with AWS (not your daily works working on a small part of it).
: Also take advantage of AWS internal metrics/on-calls to know what's
: happening with the startups, which ones are good.
: 2) Stepping stone to other big companies such as FLG or even M$ --> try join
: AWS's "front-line" products, i.e. those newly developing ones. Kinesis is
: pretty good. EMR probably is still okay. There might be other unpublished
: ones but if you can join you can use that to step to MS Azure or Google
|
