Research And Development Engineer

AI as shared language infrastructure: systems that bring disciplines together, not another private assistant.

National University of Singapore — School of Computing Full-time · On-site (Singapore) · [Contract duration: 2 years, renewable]

A funded research programme at NUS School of Computing is hiring a Research Engineer to build the core systems for a multi-year project at the intersection of knowledge graphs, taxonomies and ontologies, large language models, and human-AI interaction.

The vision: today's AI assistants are private — each person consults their own copilot, and the answers stay in their own window. Used this way, AI can quietly deepen the silos between disciplines rather than bridge them. We are building the opposite: AI as a shared language that sits between people. The biggest breakthroughs increasingly happen across disciplines, yet experts talk past each other — the same words carry different meanings, and evidence means different things across fields. Our systems help interdisciplinary teams find the right collaborators, surface and repair these hidden misunderstandings in shared spaces where everyone can see them, and turn hard-won mutual understanding into knowledge the whole team keeps. The work has direct pathways to impact in how research teams form and work together — at NUS, across Singapore's research ecosystem, and beyond — with your systems deployed to real users and your contributions feeding publications at top venues.

You will be the primary engineer turning research designs into working, evaluable systems, working closely with faculty and PhD students. Further project specifics will be shared with shortlisted candidates.

This is a builder's role: you will own substantial components end to end — from large-scale data pipelines through graph and LLM back ends to instrumented user-facing prototypes.

• Build data pipelines that ingest large document corpora and structured datasets, and extract entities, concepts, and relations at scale.
• Implement and optimise graph algorithms over large heterogeneous graphs, including clustering, partitioning, and hierarchy construction.
• Build and maintain taxonomies and ontologies over large concept collections: automated taxonomy induction from text and graph signals, ontology design and population, and alignment/matching across vocabularies.
• Develop and maintain a knowledge-graph layer (property graph and/or RDF) with validation and provenance tracking.
• Build LLM-based components: retrieval-augmented generation, structured extraction, and multi-agent pipelines with systematic evaluation.
• Prototype web-based interfaces with logging and instrumentation to support user studies.
• Set up benchmarks, baselines, and evaluation harnesses; keep pipelines reproducible and demo-ready.

Required

• Bachelor's or Master's in Computer Science or a related field.
• Strong Python engineering skills and a track record of owning systems end to end.
• Hands-on experience building LLM applications (prompting, RAG, structured output, evaluation).
• Working knowledge of graph data systems (e.g., Neo4j, RDF/SPARQL, or large-scale graph processing).
• Familiarity with taxonomies and ontologies: concept hierarchies, subsumption/is-a reasoning, and how structured vocabularies are built, validated, and used in applications.
• Solid NLP/IR fundamentals: embeddings, semantic search, entity and relation extraction.
• Able to read technical papers and implement methods from them independently.

Strong plus

• Deeper semantic-web experience: OWL modelling, SHACL validation, ontology matching/alignment tools (e.g., LogMap), reasoners, and provenance standards (PROV-O).
• Experience with automated taxonomy induction or hypernym/subsumption extraction from corpora, or hands-on work with large existing ontologies (e.g., Wikidata, UMLS/SNOMED, FIBO, or domain thesauri).
• Large-scale graph algorithms or distributed data processing (Spark, Ray, or similar).
• Front-end prototyping (React or Streamlit), especially instrumented interfaces for studies.
• Multi-agent LLM frameworks and agent orchestration.
• Prior research-lab experience, publications, or open-source research systems.
• Intention to pursue a PhD or MComp at the NUS Department of Computer Science will be an added advantage — this role offers a strong pathway into graduate study, with research experience, publications, and faculty mentorship built up before you apply.

• Work on a contrarian, meaningful thesis: AI that connects people instead of isolating them — with real deployments and measurable impact rather than throwaway prototypes.
• High-ownership role on a well-resourced, multi-year programme with a small senior team and direct faculty mentorship.
• Your work feeds directly into publications at top venues and into systems used by real users.
• An ideal launchpad for candidates planning to pursue a PhD or MComp at the NUS Department of Computer Science.
• Access to significant compute resources.
• Competitive salary commensurate with experience: [salary range].

Send your CV, a brief note on a system you have built end to end, and links to code or publications to [email protected] with the subject line "Research Engineer — Shared Language Infrastructure". Applications reviewed on a rolling basis.

• NUS is an equal opportunity employer. Only shortlisted candidates will be notified.