Machine Learning Engineer at Cosine.sh - Pioneering AI for software engineering

Cosine.sh is seeking a Machine Learning Engineer to lead the charge in developing state-of-the-art software engineering LLMs. You'll own large-scale training of our Lumen Enterprise models, leveraging supervised fine-tuning and reinforcement learning to push performance on complex coding tasks. If you thrive working close to the metal with PyTorch and distributed training, and want to make a tangible impact on how engineers build software, this pioneering role in London is for you.

Overview

Role overview

Join Cosine.sh as a Machine Learning Engineer to lead large-scale training of Lumen Enterprise models, our open-source-based software engineering LLMs. You'll drive state-of-the-art performance using supervised fine-tuning and reinforcement learning, working close to the metal with PyTorch and distributed training to make big models effective in real-world software tasks.

Company

Cosine.sh

Software

Cosine is an autonomous, on-premise coding agent post-trained on human reasoning data to deliver unmatched software-engineering accuracy, security, and speed for regulated enterprises.

Responsibilities

What you will do

Take open-source base models (code + general LLMs) and turn them into high-performance Lumen Enterprise SWE agents via SFT and RL.
Design and run large-scale training experiments on multi-node GPU clusters, including long-context training and MoE-style architectures.
Build and iterate on large-scale RL loops where models write code, run tests or tools, and get rewarded (or penalized) accordingly.

Candidate profile

Who this is a fit for

Possesses strong experience training deep learning models in production, typically 3–5+ years as an ML engineer / applied scientist.
Demonstrates deep proficiency with PyTorch and its primitives, including hands-on experience with torch.distributed for scaling training.
Has experience training large sequence models or LLMs (≥70B parameters) end-to-end on multi-GPU setups, understanding practical issues.

What makes it remarkable

Why this role is remarkable

Your work directly shapes the next generations of Lumen Enterprise SWE models that engineers use every day, offering significant impact.
You’ll operate at real scale, working with large, modern open-source models, long context lengths, and multi-node training runs.
This is a full-stack ML engineering role, spanning custom PyTorch code, distributed systems, data curation, RL design, and MLOps.

Meet Jack