Job Description Summary

Job Description

The Knowledge Discovery & High Assurance Systems organization is seeking PhD student interns to advance research and prototyping at the intersection of large language models (LLMs), natural language processing (NLP), vision-language models (VLMs), and knowledge graphs. Interns will help design, implement, and evaluate models and systems that combine unstructured text, visual data, and structured knowledge to deliver trustworthy, high-impact AI capabilities for the aerospace industry.
 

Role Overview

• Contribute to research on multimodal reasoning, retrieval-augmented generation, and grounding LLMs with knowledge graphs

• Build end-to-end prototypes: data pipelines, model training/fine-tuning, evaluation, and front-end demos

• Collaborate with researchers and engineers to translate research into production-ready components

Key Responsibilities

• Design, implement, and evaluate LLM/NLP/VLM models (pretraining, fine-tuning, instruction tuning as appropriate)

• Develop and integrate knowledge graphs for reasoning, validation, and explainability (entity/relation extraction, ontology alignment, graph embeddings)

• Build retrieval-augmented and tool-augmented systems (RAG, function/tool calling, agents) with robust evaluation

• Create data pipelines for multimodal datasets (curation, augmentation, labeling) and enforce data quality standards

• Implement benchmarking and evaluation frameworks (task metrics, bias/robustness checks, human-in-the-loop reviews)

• Develop interactive demos and internal tools using modern web stacks (React, Node.js) to showcase capabilities

• Write clean, well-tested code; document methods and results; contribute to publications, tech reports, and research artifacts

Required Qualifications

• Currently enrollment in a PhD program in Computer Science, Electrical Engineering, AI/ML, or related field

• Permanent legal authorization to work in the US is required. This role is restricted to U.S. persons. GE will require proof of status prior to employment.

• Must be willing to work from the GE Aerospace Research office in Niskayuna, NY.

• Minimum GPA 3.0 / 4.0 scale
  ​

Desired Characteristics

• Research experience in at least one of: LLMs, NLP, vision-language/multimodal learning, or knowledge graphs

• Strong programming skills in Python and JavaScript

• Hands-on experience with modern AI/ML frameworks and tooling: PyTorch, TensorFlow/JAX, Hugging Face ecosystem (Transformers, Datasets, PEFT), OpenAI/Anthropic/Vertex AI SDKs

• Vector databases and retrieval tooling, LangChain/LlamaIndex

• Data and experiment, Docker, Git

• Front-end and back-end development skills: React and Node.js

• Solid understanding of ML fundamentals: optimization, evaluation metrics, overfitting/regularization, data leakage prevention

• Publications or preprints in top venues (e.g., NeurIPS, ICML, ICLR, ACL, EMNLP, CVPR, ICCV)

​​Experience with:

• Multimodal fusion and grounding; VLMs (e.g., LLaVA)

• Knowledge graph construction and query (RDF/OWL, SPARQL, Neo4j), graph learning (PyG, DGL)

• Prompt engineering, function/tool use, and agentic workflows

• RAG evaluation (faithfulness, grounding, hallucination detection) and safety alignment

• MLOps for research: reproducible pipelines, CI/CD, containerization

• Familiarity with security, privacy, and compliance considerations for AI systems

Success Indicators:

• Demonstrated improvement on benchmark tasks

• High-quality prototypes and demos

• Clear, reproducible research artifacts (papers, code, datasets, and documentation)

• Effective collaboration and communication across research and engineering

The pay range for this position is $1,000-$2,000 USD weekly. The specific weekly rate offered to a candidate may be influenced by a variety of factors including the candidate’s experience, their education, and the work location.  GE provides travel reimbursement and housing stipend for qualified interns.