Master Thesis in Adaptive Architecture for Human-AI Annotation Systems
Core
Design, implement, and evaluate a confidence-aware adaptive architecture for human-AI annotation systems to optimize review depth based on model confidence signals.
Role type
Master's thesis researcher (software engineering)
Builds
Modular architecture with interfaces for confidence providers, decision policies, and annotator feedback within an existing annotation tool.
Domain
Autonomous driving / Software Engineering / Human-AI interaction
Deliverable
production ML models
Required skills
Python programming, modular software design, interface design, automated testing, version control, empirical software engineering
Preferred skills
AI-assisted systems, machine learning, computer vision, human-computer interaction
Technologies
Python
Responsibilities
Design and integrate modular architecture; implement confidence-provider components; develop automated tests; benchmark adaptive architecture against existing workflows; conduct user studies on verification behavior.
Seniority
Master's student (final year)