Project Details

[Return to Previous Page]

Secure Local AI Platform for Protected Information Environments

Company: PSU College of Engineering

Major(s):
Primary: CMPSC

Non-Disclosure Agreement: NO

Intellectual Property: NO

Organizations increasingly need AI-assisted learning, knowledge management, and engineering support capabilities, but many cannot use public AI services because the information involved is sensitive, proprietary, export-controlled, security-related, or otherwise restricted from external disclosure. Modern AI tools have the potential to significantly improve education, training, knowledge management, engineering analysis, and technical decision support. However, many organizations cannot utilize public AI services because their information may be sensitive, proprietary, export-controlled, security-related, privacy-protected, or otherwise restricted from external disclosure. This project seeks to develop and evaluate a locally deployed AI platform capable of supporting these activities while maintaining organizational control of data, with robust data integrity, and complying with applicable security, privacy, intellectual property, and regulatory requirements. While this project will use representative test data, actual use cases could include: • Critical infrastructure (e.g., nuclear facilities) • Healthcare systems • Defense and national security • Export-controlled environments (ITAR/EAR) • Proprietary industrial operations • Research laboratories Project deliverables can include: • Local LLM deployment and benchmarking • Evaluation of open-source models for different use cases • Secure document ingestion and retrieval • Role-based access concepts • Local RAG architecture • Hardware recommendations • User experience design