Research Fellow: AI, Multi-Agent & RAG

Nanyang Technological University, Singapore (NTU Singapore); Singapore; On-site; SGD 70,000 - 120,000; Full time

Job summary

ATMRI at NTU Singapore seeks a Research Fellow to lead the Digital ATCO Assistant project, focusing on knowledge, retrieval, reasoning, and agent orchestration with LLMs and multi-agent architectures. You will design, implement, and evaluate agentic AI systems and tool calling capabilities within a prototype environment.

The role requires PhD or substantial applied research in AI, with strong Python experience and familiarity with PyTorch, vector databases, and knowledge graphs.

Qualifications

• PhD or Master’s with substantial applied research in AI/CS/ML/NLP.
• Hands-on experience with LLMs, Agentic AI, and multi-agent architectures.
• Experience with retrieval augmented generation and knowledge representations.
• Strong Python skills and experience with modern LLM frameworks.

Responsibilities

• Design agentic AI architectures and multi-agent task decomposition.
• Develop tool calling capabilities for tool/APIs and handle outputs/errors.
• Build RAG pipelines across AIPs, NOTAMs, and procedures using vector DBs.
• Design benchmarks to evaluate performance, grounding, and hallucination.
• Expose capabilities via APIs/microservices for prototype integration.
• Translate use cases into research datasets and test scenarios.
• Contribute to reports, proposals, and publications with partners.

Skills

LLM

Agentic AI

Multi-agent systems

Tool calling

Prompt engineering

Context engineering

RAG

Vector databases

Semantic retrieval

Knowledge graphs

Python

PyTorch

LLM frameworks

APIs

Microservices

LoRA/PEFT

In-context learning

Graph RAG

Temporal RAG

Symbolic-neural reasoning

On-prem deployment

Inference optimization

Collaboration skills

Education

PhD in AI/CS/ML/NLP

Master's with applied research

Tools

Vector databases

Knowledge graphs

APIs

Microservices

PyTorch

Job description

ATMRI at NTU Singapore seeks a Research Fellow to lead the Digital ATCO Assistant project, focusing on knowledge, retrieval, reasoning, and agent orchestration with LLMs and multi-agent architectures. You will design, implement, and evaluate agentic AI systems and tool calling capabilities within a prototype environment.

The role requires PhD or substantial applied research in AI, with strong Python experience and familiarity with PyTorch, vector databases, and knowledge graphs.