Senior Data Scientist LLM Fine-Tuning & Recommendation Systems || San Jose, CA (2days / week)

Seeking a hands-on Senior Data Scientist with strong expertise in LLM Fine-Tuning, Recommendation Systems, and Ranking Models. The ideal candidate must have proven experience training, adapting, and evaluating transformer-based models using large-scale datasets, with deep knowledge of recommendation science and model optimization on AWS., * Build large-scale training datasets from structured and unstructured data.

• Fine-tune foundation models using LoRA, QLoRA, PEFT, and SFT.
• Develop recommendation, personalization, and ranking models.
• Design experiments and evaluate model performance using NDCG, MRR, MAP, and other ranking metrics.
• Perform error analysis, model benchmarking, and data quality assessments.
• Train and optimize models on Amazon SageMaker and distributed environments.
• Collaborate with science teams to improve ranking, reasoning, and recommendation quality.

• Strong LLM Fine-Tuning experience (LoRA, QLoRA, PEFT, SFT) required
• Hands-on experience working with model weights, transformer training loops, and model adaptation
• PyTorch and Hugging Face Transformers
• RLHF, DPO, PPO, or GRPO
• Recommendation Systems, Ranking Models, Personalization
• NDCG, MRR, MAP and model evaluation frameworks
• Amazon SageMaker
• PySpark, Spark, Hive, Hadoop
• Statistical Modeling & Machine Learning

Preferred Skills

• SageMaker HyperPod
• vLLM, Quantization, Distillation
• AWS Trainium / Inferentia
• Experience building large-scale recommender systems