We are looking for members of technical staff specializing in ML. We’re particularly interested in self-motivated researchers and engineers who want to meaningfully contribute towards training powerful models, no matter whether that means working on low level GPU optimizations or new optimization theory.
Relevant Skills (not all are necessary):
Work on large language models at a lab (e.g. OpenAI, Google, Mistral, Z.ai, Qwen, Deepseek, Ai2, or academic).
Experience with pretraining language models and large scale AI infrastructure, can include any of the following:
Contributions to community initiatives such as NanoGPT Speedrun, Marin, etc.
Understanding of different types of model parallelism.
Software/hardware codesign to maximize training throughput.
Experience with monitoring and maintaining large scale training runs.
Academic works (e.g. papers on optimization, data, etc).
Experience with post-training language models
Work on reinforcement learning for language models (environments, infrastructure, training).
Academic or personal work on instruction data curation, tool use, or generally post-training related tasks.
Experience with inference/systems optimization
Contributions to vLLM, SGLang, Dynamo, MegaKernels, etc
Strong systems level understand of these frameworks and how to optimize for batching, KV cache pressure, long context, etc
Experience with low level kernel design + DSLs
CUDA, C++, CuTE, Triton, PTX, TileLang, etc
Developing AI foundation models that can run efficiently on any device. They're working on a new large language model architecture to make these models more efficient. Their big goal is to create a top-performing model, similar to those from OpenAI or Google, that can run entirely on a phone locally. They're focused on making AI that's not just powerful but also super efficient.
Salary
$200,000 - $350,000
Location
San Francisco
Total raised
$9.0M
Last stage
Pre-seed
Investors
Shayaan Emran
Edward Zhang
No applications, no recruiter spam. Just the intro.
A few questions to make sure this role is the right shape for you. Two minutes.
I write the intro, send it to the founder, and handle the back-and-forth.
If they’re a yes, I book the chat. You show up — that’s the whole job-hunt.