arrow
Return

Engineering Agentic Systems: Toward Verification-First AIware Engineering

delete2026-09-07
delete0
PRE
AI
Q
Qinghua Lu
A
Ahmed E. Hassan
D
David Lo
T
Thomas Zimmermann
DOI:10.1109/ms.2026.3697911delete
deleteOriginal
deleteOriginal request for help
deleteShare
deleteSave
Abstract

Abstract

En 中文
Agentic systems are transforming software engineering from a code-centric discipline toward AIware engineering, where intelligence is deliberately allocated across foundation models, on-demand generated software, and external tools. The central challenge is no longer generating behaviours, but verifying them at scale. Verification-Driven AIware Engineering addresses this challenge by treating verifiability as a first-class design objective, deliberately decomposing problems to enable scalable verification.
Keywords:
Special issues and sections
Agentic AI
Formal verification
Feedback loop
Software engineering
Codes
Encoding
Scalability
Human in the loop
Multi-agent systems

Journal

IEEE Software cover
IEEE Software
IF:
3
Papers:
3.2K
Citations:
3.6K

Organization

U
University of California
Scholars:
9.7K
Papers: 3.6K
Citations: 8.3W
Q
queen's university
Scholars:
1.0K
Papers: 480
Citations: 0
C
csiro
Scholars:
1.0K
Papers: 509
Citations: 0
S
singapore management university
Scholars:
371
Papers: 278
Citations: 0
researcher View more organizations
Cited Papers

Cited Papers

No cited papers available