LLM Fine-Tuning and Alignment Engineer - Content Developers

Jalasoft

15h•Remote

About The Position

Jala University is seeking content developers for its Master in AI Systems Engineering program. This role involves creating comprehensive educational materials for LLM fine-tuning and alignment, including SFT, PEFT/LoRA, DPO/KTO, verifier-driven RL, domain-data preparation, preference optimization, and alignment-auditing methodology. The output should include module content, labs, an evaluation rubric, and an instructor guide, all produced to a production standard.

Requirements

5+ years shipping production software
2+ years in production AI/ML
Production SFT, PEFT and preference-optimization experience on real domain data
An alignment-audit track record
Comfortable at multi-GPU scale
Disciplined about run lineage
Ability to build the artifact personally, to production standard
Public writing samples showing technical explanation (docs, workshop material, book chapter, or open-source project known for its docs), including ability to write to publication standard
Reproducibility discipline (pinned dependencies, containers, seeded runs, documented decoding parameters)
Professional written English

Nice To Haves

Experience with Axolotl, Unsloth, PEFT, TRL, Hugging Face Inference Endpoints, Modal (H100 / A100 80GB), Weights & Biases, OpenRouter.

Responsibilities

Develop educational content for LLM fine-tuning and alignment, covering SFT, PEFT/LoRA, DPO/KTO, verifier-driven RL, domain-data preparation, preference optimization, and alignment-auditing methodology.
Create module content, labs, an evaluation rubric, and an instructor guide.
Ensure all developed artifacts meet production standards.
Write public technical explanations (docs, workshop material, book chapter, or open-source project known for its docs) to a publication standard.
Implement reproducibility discipline (pinned dependencies, containers, seeded runs, documented decoding parameters).