Machine Learning Engineer

Second Talent; Malaysia; Remote; MYR 250,000 - 300,000; Full time

Job summary

A tech company in Malaysia seeks an Environment Engineer (ML) for building simulation environments for machine learning coding agents. The role involves developing testing infrastructure, implementing algorithmic improvements, and requires extensive experience with ML frameworks like PyTorch or Jax. The interview process is designed to be quick, often leading to an offer within a week.

Qualifications

• Multiple years of experience in machine learning environments.
• Familiarity with testing infrastructure and automation for ML.
• Proven ability to implement algorithmic improvements.

Responsibilities

• Build simulation environments for ML coding agents.
• Develop end-to-end tests for machine learning models.
• Research and implement algorithmic improvements for model training.

Skills

Experience with ML frameworks like PyTorch or Jax

Ability to quickly familiarize with new technologies

Experience with large models and datasets

Job description

Member of Technical Staff - Environments (ML)

Remote, Full-Time

As an Environment Engineer (ML), you will build on top of our core platform to create the simulation environments in which frontier coding agents learn to be proficient in Machine Learning. You will build testing infrastructure on top of large codebases to grade the solutions of coding agents on real world ML engineering tasks and build automations to speed up parts of testing and task creation.

Here are some things you might work on:

• Write an end-to-end test for an RL framework that trains a small model to ensure that the reward of our model is going up as expected or numerical instabilities occur
• Find research papers proposing algorithmic improvements for model training, implement them and write extensive tests to ensure that we can test whether models are able to reimplement them

This role will require you to constantly work with new codebases; hence, you will need to be able to quickly get familiar with new technologies. The ideal candidate has multiple years of experience working with ML frameworks like PyTorch or Jax and has trained large models on large datasets using large amounts of GPUs.

Interview Process

Our interview process is designed to be quick. We usually have 2-3 technical interviews before making an offer. You can expect to have an offer within seven days in most cases.