AI Agent Architect: Tool-Use & LLM Integration (Flexible Hours)

Barcelona Supercomputing Center; Barcelona; On-site; EUR 73,000 - 95,000; Full time

Benefits offered by this job

Restaurant tickets

Private health insurance

Relocation support

Flexible working hours

Job summary

The Barcelona Supercomputing Center (BSC-CNS) invites applications for an AI Agent & Tool-Use Engineer (RE2) to advance agentic workflows, multi-step reasoning, and tool integration within a premier HPC environment in Barcelona.

You will collaborate with the Language Modeling Team on fine-tuning and post-training of large language models, explore Model Context Protocol interfaces, and build evaluation pipelines for reliable model-tool interactions.

Qualifications

• The candidate should have strong experience in fine-tuning and post-training of large language models.
• Hands-on experience with agentic AI patterns and tool-use integration.
• Proficiency in Python and deep learning frameworks such as PyTorch.
• Experience with post-training frameworks like TRL, FastChat, NeMo-RL.

Responsibilities

• Implement and test agentic workflows including multi-step reasoning and tool-use capabilities under supervision.
• Develop and refine Function Calling interfaces and connectors using MCP to enhance model interactivity.
• Execute post-training experiments to improve agentic performance of models.
• Build and maintain evaluation pipelines for model-tool interactions.
• Collaborate with the data team to curate instruction datasets for task-oriented dialogue.

Skills

Fine-tuning LLMs

Post-training LLMs

Agentic AI patterns

Python

PyTorch

MCP API integration

Model Context Protocol

English proficiency

Spanish knowledge

Education

Degree in Computer Science, Mathematics or related

Tools

TRL

FastChat

NeMo-RL

Job description

The Barcelona Supercomputing Center (BSC-CNS) invites applications for an AI Agent & Tool-Use Engineer (RE2) to advance agentic workflows, multi-step reasoning, and tool integration within a premier HPC environment in Barcelona.

You will collaborate with the Language Modeling Team on fine-tuning and post-training of large language models, explore Model Context Protocol interfaces, and build evaluation pipelines for reliable model-tool interactions.