The National Institute of Education invites suitable applications for the position of Research Assistant on a 12-month contract at the English Language and Literature department.
Project Title: Train Your Dragon: Fine-tuning Large Language Models to Create Agentic AI for Automatic Assessment Generation
Project Introduction:
This position involves two closely related research strands at the intersection of language assessment and artificial intelligence. The first focuses on developing agentic AI systems to support the automated construction and evaluation of language tests. This includes fine-tuning large language models (LLMs) for tasks such as item generation, test assembly, and automated essay scoring, with an emphasis on validity, reliability, and methodological transparency.
The second strand centers on advancing LLM-based spoken dialog systems integrated with sensor technologies, including neuroimaging and eye-tracking. Building on our existing in-house platform, the research assistant will extend system capabilities to address emerging research questions related to cognitive load, interaction dynamics, and adaptive dialogue management. This role requires both technical development and empirical research, contributing to innovative approaches in technology-enhanced language assessment.
A. Required background and skills
- Bachelor’s degree in computer science, artificial intelligence, data analytics, computational psychology, or a closely related field
- Demonstrated experience with large language models (LLMs), including fine-tuning, APIs, or related methods for text generation or evaluation tasks
- Programming proficiency in Python (knowledge of R is an advantage)
- Experience developing agentic AI systems using LLM APIs/SDKs, including tool use, orchestration, and workflow design
- Proficiency in one or more of the following areas: natural language processing (NLP) and/or speech technologies (e.g., ASR, TTS, dialog systems)
- Reliable background in quantitative and computational methods
- Working understanding of experimental research design and data analysis
B. General competencies
- Strong motivation for interdisciplinary research and ability to work collaboratively in a team environment
- Strong teamwork skills, openness to innovative ideas, and intellectual curiosity
- Resilience and persistence in tackling complex research and development challenges
- Excellent written and verbal communication skills in English
C. Preferred
- Experience with automated item generation, automated essay scoring, or educational measurement applications of AI
- Familiarity with spoken dialog systems or conversational AI development
- Familiarity with sensor-based research methods (e.g., EEG, fNIRS, eye tracking)—training will be provided
- Familiarity with neurofeedback and human–computer interaction (HCI) methods
- Experience working with multimodal data (e.g., speech, text, physiological signals)
- Prior involvement in interdisciplinary projects combining AI, language, and cognitive science
- Experience contributing to academic publications or conference presentations
Responsibilities:
- Develop and fine-tune LLMs for language assessment applications, including automated item generation, test assembly, and automated essay scoring
- Design and implement agentic AI systems using LLM APIs/SDKs, including tool integration, workflow orchestration, and iterative model improvement
- Contribute to the development and extension of in-house spoken dialog systems for research and experimental use
- Integrate and manage multimodal data streams (e.g., speech, text, neuroimaging, eye tracking) within dialog system architectures
- Support the design and implementation of experimental studies involving human-AI interaction and sensor-based data collection
- Analyze and interpret quantitative and multimodal datasets using appropriate statistical and computational methods
- Collaborate with interdisciplinary team members across AI, language assessment, and cognitive science domains
- Contribute to the preparation of research outputs, including conference papers, journal articles, and technical reports
- Document code, models, and experimental procedures to ensure reproducibility and scalability
- Assist in maintaining and improving research infrastructure, including codebases, APIs/SDK integrations, and experimental platforms
- Any other duties as assigned by the Principal Investigator
Application
Applicants (external and internal) will apply via Workday. We regret that only shortlisted candidates will be notified.
Closing Date
Closing date for advertisements will be set to 14 calendar days from date of posting.
Hiring Institution: NIE