Undergraduate Research Assistant (AI for Scientific Discovery)

Project Overview

We are developing next-generation AI systems that can move beyond summarizing scientific knowledge to generating new, testable hypotheses from complex, multimodal data. The project integrates large language models, knowledge graphs, simulation tools, and experimental data into an agentic reasoning framework for scientific discovery.

The application domain includes materials science (e.g., corrosion-resistant alloys), but the core focus is on AI, machine learning, and scientific reasoning systems.

Responsibilities

• Develop and test components of an AI-driven scientific reasoning system
• Work with large language models (LLMs) and tool-calling frameworks
• Build and query knowledge graphs (Neo4j, Cypher)
• Process and integrate multimodal data (text, images, simulation outputs)
• Assist in implementing machine learning models (e.g., CNNs, GNNs, embeddings)
• Contribute to software pipelines for data extraction, modeling, and analysis

Preferred Qualifications

We strongly encourage applicants with experience in software and AI, including:

• Python programming (required)
• Familiarity with machine learning (PyTorch, TensorFlow, or similar)
• Experience with data processing, APIs, or backend development
• Interest in AI for science, reasoning systems, or autonomous agents

Nice to have (but not required):

• Experience with LLMs, knowledge graphs, or scientific computing
• Background in materials science, physics, chemistry, or engineering
• Experience with image analysis or simulation data

Who Should Apply

• Undergraduate students in Computer Science, Data Science, Engineering, Physics, Chemistry, or related fields
• Students interested in working at the intersection of AI and scientific discovery
• Motivated learners who want hands-on experience in a cutting-edge research project

What You Will Gain

• Experience working on state-of-the-art AI systems for science
• Exposure to multimodal machine learning and agent-based reasoning
• Opportunity to contribute to publications and NSF/DOE-funded research
• Mentorship in both AI/ML and scientific computing

How to Apply

Please send:

• Resume
• Brief statement of interest (1–2 paragraphs)
• (Optional) GitHub or project portfolio

Please email at mahataa@merrimack.edu

Pay: $25.11 - $30.25 per hour

Work Location: Hybrid remote in North Andover, MA 01845