Senior Applied Scientist: LLM & Conversational AI

Amazon; Torino; In loco; EUR 94.000 - 115.000; Tempo pieno

Descrizione del lavoro

Amazon Development Centre Italy SRL in Torino seeks a Senior Applied Scientist for the Alexa AI team to define the science roadmap for state-of-the-art conversational AI powered by large language models. You will lead LLM fine-tuning, alignment, and agentic architectures for production deployment.

Working at the intersection of research and production, you will build evaluation frameworks, automated training pipelines, and conduct rigorous experiments, collaborating with engineering and product

Competenze

• Several years of building ML models for business applications.
• PhD or Master's degree in a relevant field.
• Proficiency in Java, C++, Python.
• Experience with neural deep learning methods.

Mansioni

• Define and drive the science roadmap for conversational AI capabilities powered by large language models.
• Design, implement, and evaluate LLM fine-tuning, alignment (RLHF, DPO, RLVR), and distillation for production deployment.
• Architect agentic systems that work reliably at scale.
• Develop evaluation frameworks and methodologies beyond standard benchmarks.
• Translate research advances into customer-facing products and collaborate with engineering and cross-functional teams.
• Own end-to-end delivery of complex research initiatives from problem formulation to production deployment.
• Advance the team's scientific reputation through publications and IP.
• Publish findings and contribute to patents.

Conoscenze

Java

C++

Python

TensorFlow

PyTorch

Formazione

PhD or Master's degree

Descrizione del lavoro

Amazon Development Centre Italy SRL in Torino seeks a Senior Applied Scientist for the Alexa AI team to define the science roadmap for state-of-the-art conversational AI powered by large language models. You will lead LLM fine-tuning, alignment, and agentic architectures for production deployment.

Working at the intersection of research and production, you will build evaluation frameworks, automated training pipelines, and conduct rigorous experiments, collaborating with engineering and product