Huawei Technologies Canada Co., Ltd.Hiring

Senior AI System Engineer, LLM Training System

  • USD 127,000 to 225,000 per year
  • Vancouver, BC
  • Senior
  • Posted yesterday
Apply with Keywordise AI

or apply from Claude / ChatGPT / Muse

In one line

Senior systems role tuning LLM training on Ascend NPUs, writing high-performance kernels and scaling distributed pipelines; permanent and on-site with a lab team.

About the role

Suits an engineer who has shipped accelerator operators and distributed training systems and wants to work close to hardware. Work covers kernel design, parallelism tuning and numerical parity between training and serving.

At a glance

Pay
USD 127,000 to 225,000 per year
Location
Vancouver, BC
Seniority
Senior
Skills
AscendC, Triton, TileLang, Python, PyTorch, Megatron-LM, FP8
Posted
3 Nov 2026
Last checked
3 Nov 2026

What you will do

  • Design and tune core operators such as FlashAttention variants, FP8 GEMM and fused RMSNorm in AscendC
  • Scale distributed pre-training and RL post-training pipelines with MindSpeed and upstream frameworks
  • Profile on Ascend hardware with msprof, msprobe and PyTorch Profiler to remove stalls and communication overhead
  • Contribute operators and fixes to open-source Ascend and AI acceleration projects

What they look for

  • Degree in computer science, computer engineering, electrical engineering or a related field
  • 3+ years in HPC, AI accelerator operator work or distributed large model training
  • Hands-on AscendC or CANN experience, or CUDA C/C++ with the ability to move to AscendC
  • Interest or experience with Triton or TileLang for accelerator backends
  • Strong C/C++ and modern Python, including PyTorch custom C++ or CANN extensions

Nice to have

  • Practical experience with Megatron-LM, PyTorch FSDP, veRL, vLLM or DeepSpeed

How Keywordise applies for you

  1. 1Upload your CVAdd your CV once. It takes about a minute.
  2. 2Tailored to this jobKeywordise writes a CV and a cover letter for this job from your real experience.
  3. 3The form fills itselfName, email, experience and screening questions, field by field.
  4. 4Application sentYou get a receipt and follow it in your dashboard.

Apply with Keywordise AI7-day free trial, no card.

Apply from your AI assistant

Connect Keywordise to your assistant once, then send it the prompt. It applies for you.

ClaudeConnectorChatGPTConnectorMuseby Meta

Claude

  1. In Claude, open Settings, then Connectors, and choose Add custom connector.
  2. Name it Keywordise and paste the connector URL.
  3. Click Connect, log in to Keywordise and approve. Then send the prompt in a new chat.
Connector URL
https://keywordise.com/mcp

Disconnect any time in your Keywordise settings.

ChatGPT

  1. In ChatGPT, open Settings, then Connectors. On plans that need it, turn on developer mode first.
  2. Add a new connector with the connector URL and choose OAuth. Leave the client ID and secret empty.
  3. Log in to Keywordise and approve. Then send the prompt in a new chat.
Connector URL
https://keywordise.com/mcp

Disconnect any time in your Keywordise settings.

Muse

  1. Open Muse (muse.ai or the app) and send it the message below.
  2. If Muse asks how to sign in, answer OAuth. Open the sign-in link it sends, log in to Keywordise and approve.
  3. Send the prompt. Muse asks before it sends anything.
Message for Muse
Add a custom connector called Keywordise. It is a remote MCP server at https://keywordise.com/mcp and uses OAuth. Give me the sign-in link.

Muse is available in the US only for now.

Prompt
Apply to Keywordise job jcg5eebm2e
How to connect

Questions about this job

Where is this job based?

The posting lists Senior AI System Engineer, LLM Training System as a role based in Vancouver, BC, Canada.

What does this job pay?

The posting states USD 127,000 to 225,000 per year.

What experience is needed?

It is a senior role. The posting asks for: degree in computer science, computer engineering, electrical engineering or a related field; 3+ years in HPC, AI accelerator operator work or distributed large model training; hands-on AscendC or CANN experience, or CUDA C/C++ with the ability to move to AscendC. Skills it names include AscendC, Triton, TileLang, Python.

How do I apply with Keywordise AI?

Click Apply with Keywordise AI, create an account and upload your CV. Your CV and cover letter are written for this job from what you have really done, then Keywordise applies for you. Trial: 7 days, no card.

More IT and software jobs in Vancouver, BC

All IT and software jobs in Vancouver, BC
Huawei Technologies Canada Co., Ltd.Vancouver, BC
Apply with Keywordise AI