Remote AI Engineer: LLM Fine-Tuning for Production

Workana LLC; United States; Remote; USD 70,000 - 120,000; Part time

Benefits offered by this job

Fully remote role

Real AI product work in legal/health

High ownership and autonomy

Performance-based incentives

Opportunity to transition to full-time

Job summary

Medxprts.ai is hiring an AI Engineer - LLM Fine-Tuning to hands-on train and fine-tune open-weight models, build training infra, and create robust evaluation and feedback pipelines. You will work with PyTorch, Hugging Face tooling and cloud providers to ship models into production.

Role requires collaboration with engineering teams, experience with LoRA/QLoRA and SFT, and overlap with U.S. time zones. Remote work with potential full-time transition is offered.

Qualifications

• Hands-on experience fine-tuning LLMs, open-weight models and shipped example.
• Experience building training infra with PyTorch & Hugging Face toolings.
• Dataset engineering for messy unstructured docs; deduplication and train/eval splits.
• Strong evaluation discipline with human-calibrated or regression testing.
• Experience with preference/feedback learning methods and model updates.

Responsibilities

• Design, run LLM fine-tuning experiments (LoRA/QLoRA and SFT) on open-weight models.
• Build and maintain training infra across multi-GPU setups (DeepSpeed/FSDP).
• Engineer domain datasets from long PDFs and medical/legal records.
• Create evaluation harnesses and automated drift monitoring.
• Integrate feedback loops (DPO/RLHF-style) into retraining pipelines.
• Collaborate with backend/frontend teams to deploy models in production.

Skills

LLM fine-tuning

Dataset engineering

English communication

Tools

PyTorch

Hugging Face (Transformers)

PEFT

TRL/Axolotl

DeepSpeed/FSDP

AWS/GCP

Docker

Kubernetes

Job description

Medxprts.ai is hiring an AI Engineer - LLM Fine-Tuning to hands-on train and fine-tune open-weight models, build training infra, and create robust evaluation and feedback pipelines. You will work with PyTorch, Hugging Face tooling and cloud providers to ship models into production.

Role requires collaboration with engineering teams, experience with LoRA/QLoRA and SFT, and overlap with U.S. time zones. Remote work with potential full-time transition is offered.