"Student Research Worker"

Job Details

• Posted: 18-Jan-26
• Location: New York, New York
• Type: Part-time
• Categories: Staff/Administrative
• Job Type: Short Term Casual
• Hours Per Week: 10
• Salary Range: $23.00/Hr - $23.00/Hr

Position Summary

This position supports an IRB-approved research project within the SpeechCARE Lab at Columbia University Irving Medical Center, focused on developing speech-based artificial intelligence and large language model (LLM) methods for early detection of cognitive impairment. The student will contribute to cutting-edge biomedical AI research using real-world patient-clinician communication data, working closely with the PI and an interdisciplinary research team.

Responsibilities

• Participate in regular research meetings (typically twice per week with the PI and team)
• Work with audio and transcript data in secure environments
• Using natural language processing, large language models, and speech analysis to conduct research on building algorithms for early detection of Alzheimer’s disease based on audio-recorded patient-clinician data.
• Contribute to prompt design, LLM fine-tuning, and model evaluation
• Assist with documentation, analysis, and reproducible research code

Minimum Qualifications

• Bachelor's Degree
• Strong Python programming skills
• Strong background in NLP, with completion of relevant graduate-level coursework
• Experience working with transformer-based models
• Familiarity with prompt writing, LLM fine-tuning concepts, and GPT-style APIs
• Experience with PyTorch and/or Hugging Face is strongly preferred
• Familiarity in speech analysis is a plus (training will be provided by the research team)

Prior to accessing data, selected students must complete required CUIMC research compliance training (e.g., CITI Human Subjects Research and HIPAA) and be added to the IRB protocol.

Preferred Qualifications

• Familiarity with speech analysis or audio processing (training provided by the research team)
• Prior experience in biomedical, clinical, or health-related AI research