Federal Project Grant Award Summary

The University of Massachusetts received a $396,112 Project Grant from the National Science Foundation (NSF) Division of Information and Intelligent Systems under the Computer and Information Science and Engineering (CISE) program (CFDA 47.070) for the period of July 1, 2026 through June 30, 2031. This CAREER award supports the development of formal specifications as foundational infrastructure for aligned and automated software engineering. The project delivers three primary products: (1) a large-scale repository dataset linking natural language documentation, formal behavioral specifications, and code artifacts with an automated evaluation pipeline for reproducible assessment; (2) automated inference methods that derive specifications from documentation and program structure, propagate them across code dependencies, and maintain semantic alignment through software version changes; and (3) specification-driven development tools that leverage formal specifications as semantic constraints to guide artificial intelligence-based code generation and detect behavioral inconsistencies early in the software development lifecycle.

The project addresses critical challenges in autonomous AI-driven software development by making formal specifications practical and accessible to practitioners. By integrating data infrastructure, inference algorithms, and alignment-oriented tooling, the research strengthens theoretical and practical foundations for ensuring that software behaves as intended. The deliverables support the development of safer AI-based autonomous systems with applications in healthcare and finance, thereby reducing costly failures and increasing public trust in critical software infrastructure.