Machine Learning Engineer

MacroCoderz; Pakistan; Remote; PKR 19,364,000 - 30,429,000; Full time

Benefits offered by this job

Remote work

Job summary

MacroCoderz is seeking a Remote Machine Learning Engineer to fine-tune, evaluate, and deploy models—from classifiers to open-source LLMs—in production systems.

You will build and maintain datasets, compare models to hosted APIs on quality and cost, and ensure reproducibility of experiments while keeping latency in check. The role is fully remote and open to applicants worldwide.

Qualifications

• 2+ years of professional ML engineering.
• Strong Python and PyTorch.
• Experience fine-tuning transformer models.
• Solid grasp of evaluation methodology and overfitting.

Responsibilities

• Fine-tune and evaluate models, including open-source LLMs and embedding models.
• Build training and evaluation datasets with the AI Engineering team.
• Serve models efficiently and monitor them in production.
• Compare models against hosted APIs on quality, cost and latency.
• Keep experiments reproducible and documented.

Skills

Python

PyTorch

Transformer fine-tuning

Evaluation skills

Tools

Hugging Face

vLLM

Job description

Remote Machine Learning Engineer at MacroCoderz: fine-tune, evaluate and serve models - from classifiers to open-source LLMs - in production systems.

Not every problem should go to a large hosted model. As a Machine Learning Engineer you build the models that make our systems faster, cheaper or more private: classifiers and rerankers, fine-tuned open-source LLMs, and embedding models adapted to a client's domain.

You take a model from dataset to production endpoint, and you prove it beats the simpler option before it ships.

What you will do

• Fine-tune and evaluate models, including open-source LLMs and embedding models.
• Build training and evaluation datasets with the AI Engineering team.
• Serve models efficiently and monitor them in production.
• Compare custom models against hosted APIs on quality, cost and latency.
• Keep experiments reproducible and documented.

What we are looking for

• 2+ years of professional ML engineering.
• Strong Python and PyTorch.
• Experience fine-tuning transformer models.
• Solid grasp of evaluation methodology and overfitting.

Nice to have

• LoRA/QLoRA and the Hugging Face ecosystem.
• Model serving with vLLM, TGI or similar.
• Published research or open-source contributions.

Location

fully remote, open to applicants worldwide.