[Job Ad]Alibaba Group is hiring deep learning compiler engineers

I’m writing this email behalf of my manager in Alibaba. Our team is working on a deep learning engine called MNN. We open sourced it this May. (Github link here: https://github.com/alibaba/MNN )).

It has industry-leading blazing fast inference speed on mobile and it is tiny in size. In the internal experiments we ran, MNN outperforms other inference engines under almost all settings by about 20% - 40%. And we are trying to make it even faster and support more computational backends by exploring MLIR, XLA, TVM and the like. We’d love to hear from anyone interested in the following position:

Company: Alibaba.

Location: Hangzhou or Beijing in China.

Language: Chinese is preferred.

Job Description: You will be working on MNN with a team of enthusiastic, energetic, talented and hard-working engineers in the lovely city Hangzhou.


  1. For this role, we are interested in people with in-depth knowledge, understanding and hands-on work with MLIR, XLA, or TVM.
  2. Contributions to the open source projects like LLVM, MLIR, XLA or TVM are a great plus.
  3. Since our product is an inference engine, knowledge and experience in the architecture, design, internals of other ML systems (e.g. Tensorflow, PyTorch) are a great plus.
  4. Experience of leading a compiler team would be preferred

Level: This is a senior level position, equivalent to a great Google L4 or solid Google L5.

If you are interested in this role, visit this link and apply
If you wish to know more about this role, you can contact my manager on his LinkedIn: https://www.linkedin.com/in/huishu/ , reply to this email or contact him directly at hush.alibaba#gmail.com


Is this offer about working on an LLVM-based project? MNN does not seem to be LLVM-based or closely related, but I may have missed something.
This list policy is to restrict job offers to LLVM-related offers (per previous examples: http://lists.llvm.org/pipermail/llvm-dev/2019-April/131728.html ).


Thanks for the interest. Yes we currently don’t use LLVM/MLIR directly, but we do plan doing so in forseeable future, hence this job ad. Sorry for the confusion!