Senior Applied Scientist, Large Language Models

About the role

Patsnap is hiring a Senior Applied Scientist to build sophisticated language model features for knowledge-heavy products. You will bridge the gap between theoretical research and production deployment by managing the full lifecycle of model development, testing, and implementation.

Key facts

What you'll do

• Create and refine language model capabilities for practical business use cases.
• Enhance model performance regarding reasoning, long-context processing, data extraction, retrieval-augmented generation, and domain-specific adaptation.
• Execute post-training workflows such as supervised fine-tuning, preference optimization, knowledge distillation, and synthetic data creation.
• Establish evaluation frameworks to monitor accuracy, factuality, safety, latency, and operational costs.
• Construct pipelines for data preparation, experimentation, and model assessment.
• Investigate model errors to determine and implement corrective strategies.
• Evaluate new research to determine its utility for production environments.
• Partner with engineering to deploy and tune models at scale.
• Translate business needs into technical solutions alongside product managers.
• Help define technical standards and the internal AI roadmap.
• Mentor junior engineers and researchers on the team.

Requirements

• Master degree or PhD in Computer Science, AI, Machine Learning, NLP, or a related field, or equivalent work experience.
• Background in machine learning, NLP, or applied AI.
• Experience building or modifying large language models.
• Knowledge of Transformer architectures, training cycles, fine-tuning, and inference.
• Proficiency in at least two of the following: LLM post-training/alignment, benchmarking, retrieval-augmented generation, long-context modeling, information extraction, complex reasoning/planning, or model compression/inference optimization.
• Coding fluency in Python and experience with PyTorch.
• Ability to manage end-to-end algorithmic projects from problem definition to production delivery.
• Communication skills for cross-functional collaboration.

Nice to have

• Experience with enterprise, scientific, or technical AI applications.
• Background in distributed training, large-scale inference, or GPU optimization.
• Experience creating automated evaluation systems, data flywheels, or human-feedback loops.
• Familiarity with multimodal models, AI agents, or tool-augmented systems.
• A record of publications in major AI/NLP conferences or open-source contributions.

Skills & tools

• Python
• PyTorch
• Large Language Models
• Transformer Architectures
• Retrieval-Augmented Generation (RAG)
• Model Fine-tuning
• Inference Optimization