LLM Fine-Tuning & Research Engineer

Serus; New York (NY); Híbrido; USD 140.000 - 210.000; Jornada completa

Descripción de la vacante

Anon AI Labs, the company behind Serus, seeks an ML Engineer/Researcher to lead LLM fine-tuning and post-training for its customer-facing intelligence product. You will design training pipelines from base model selection to data prep, training, evaluation and production deployment, working with engineers across teams in New York or San Francisco.

You will evaluate open-weight models, build datasets, fine-tune for domain knowledge and complex task execution, and develop rigorous evaluations for

Formación

• Hands-on experience fine-tuning or post-training language models.
• Strong Python and practical knowledge of modern training frameworks.
• Experience preparing high-quality datasets and designing meaningful evaluations.
• The ability to diagnose model failures and run disciplined experiments.
• An understanding of inference and production deployment.
• Experience with tool-using models, security-domain data or agent training is particularly relevant.

Responsabilidades

• Evaluate suitable open-weight models and establish strong baselines.
• Build training datasets for intelligence research, tool use and agent behavior.
• Fine-tune models for domain knowledge, decision-making and task execution.
• Develop evaluations that measure performance on realistic customer workflows.
• Improve the handling of complex security material and sensitive context.
• Investigate supervised fine-tuning, preference optimization, distillation and other relevant techniques.
• Work with agent and backend engineers to deploy trained models.
• Measure quality, latency, inference cost and operational requirements.

Conocimientos

Fine-tuning LM

Python & ML frameworks

Data preparation & evaluation

Model debugging

Deployment knowledge

Tool-using models

Descripción del empleo

Anon AI Labs, the company behind Serus, seeks an ML Engineer/Researcher to lead LLM fine-tuning and post-training for its customer-facing intelligence product. You will design training pipelines from base model selection to data prep, training, evaluation and production deployment, working with engineers across teams in New York or San Francisco.

You will evaluate open-weight models, build datasets, fine-tune for domain knowledge and complex task execution, and develop rigorous evaluations for