ML Research Engineer: Retrieval & LLM Evaluation

University College Dublin; Ireland; On-site; EUR 60,000 - 90,000; Full time

Job summary

University College Dublin invites applications for a temporary post of a Research Data/ML Engineer within UCD School of Politics and International Relations, Connected Politics Lab, RINN AI UCD. The role leads retrieval optimisation, language-model evaluation, and data pipelines across European Parliament and EU bodies, developing robust experiments and scalable processing.

Reporting to the Principal Investigator, collaboration with a Senior Full-Stack Developer and a Junior Developer; closing

Qualifications

• Primary degree in computer science, data science, machine learning, statistics or a related discipline, or equivalent professional experience.
• Demonstrated research or applied experience in natural language processing and large language models.
• Experience with retrieval-augmented generation, embeddings and vector search.
• Strong Python programming skills.
• Experience designing and running evaluations of model or system performance, including benchmarking against defined accuracy targets.
• Experience building and operating data pipelines (ETL/ELT), including web scraping or API-based ingestion.
• Experience with relational databases (e.g. PostgreSQL) and search or vector stores (e.g. OpenSearch/Elasticsearch).
• Familiarity with machine-learning frameworks such as PyTorch, Hugging Face or LangChain.
• An understanding of the operational requirements for a successful research project.
• Excellent communication skills (oral, written, presentation).
• Excellent organisational and administrative skills, including a proven ability to work to deadlines.
• CANDIDATES must demonstrate awareness of equality, diversity, and inclusion (EDI).

Responsibilities

• Lead retrieval, evaluation and machine-learning research to determine accuracy and trustworthiness of EURVISTA platform outputs.
• Operate and extend the platform's live data pipelines.
• Design and optimise retrieval architecture and evaluate embedding models.
• Benchmark query-response accuracy against a target above 90% and calibrate confidence scoring.
• Prototype and validate statistical and predictive models over legislative datasets.
• Operate and extend daily automated ingestion from European Parliament, Council of the EU and European Commission sources.

Skills

Python

NLP

LLMs

RAG

Embeddings

Vector search

PostgreSQL

OpenSearch

Elasticsearch

PyTorch

HuggingFace

LangChain

ETL/ELT

Web scraping

API ingestion

Education

Bachelor's degree in CS/DS/ML/Statistics

Tools

PostgreSQL

OpenSearch

Elasticsearch

PyTorch

HuggingFace

LangChain

Job description

University College Dublin invites applications for a temporary post of a Research Data/ML Engineer within UCD School of Politics and International Relations, Connected Politics Lab, RINN AI UCD. The role leads retrieval optimisation, language-model evaluation, and data pipelines across European Parliament and EU bodies, developing robust experiments and scalable processing.

Reporting to the Principal Investigator, collaboration with a Senior Full-Stack Developer and a Junior Developer; closing