About the Researcher
Emanuel Shirbint
Emanuel Shirbint is a Data & AI Architect and independent researcher working at the intersection of governed agent systems, LLM safety and evaluation, BI/DWH architecture, clinical AI, human and medical digital twins, and institutional governance.
Research profile
Emanuel Shirbint is a Data & AI Architect and independent researcher. His work joins two threads: applied enterprise data architecture (BI/DWH, data engineering, data governance) and conceptual research on how AI systems can be made accountable — through governed agent architectures, structure-sensitive safety evaluation, governed abductive reasoning for medical and human digital twins, and the normative architecture that fixes what an AI system is for.
Professional background
His applied practice spans business-intelligence and data-warehouse architecture, data engineering, and the lineage, contracts and quality controls that trustworthy AI depends upon. This work provides the auditable data foundation referenced across his research programmes.
Research programmes
- Governed Medical Digital Twins / MS-AGIP
- Human Digital Twins and Self-Knowledge
- LLM-Agent Safety and Topological Robustness
- Moral Grounding and Institutional AI Governance
- Clinical AI and Multiple Sclerosis Research
- BI/DWH and Data Architecture
Clinical research experience
Emanuel Shirbint is a co-author on peer-reviewed clinical research conducted with the Multiple Sclerosis Center at Sheba Medical Center and collaborating institutions, including studies of COVID-19 vaccination safety and humoral response in multiple sclerosis, brain-volume mapping in primary-progressive MS, robust spine segmentation in MRI, and digital assessment of finger dexterity. These works are presented here historically and accurately, with each author position and the peer-review status stated on the record.
Data and AI architecture
Alongside clinical and conceptual research, his practice concerns the data architecture beneath accountable AI: warehouse and BI design, data contracts, lineage and quality controls that allow AI-assisted output to be audited rather than merely produced.
Publications
Peer-reviewed journal articles
- Development of an interactive biosensing application for assessing finger dexterity · 2025 · Digital Health
- Mapping brain volume change across time in primary-progressive multiple sclerosis · 2024 · Neuroradiology
- COVID-19 vaccination in patients with multiple sclerosis: Safety and humoral efficacy of the third booster dose · 2022 · Journal of the Neurological Sciences
- COVID-19 vaccination in patients with multiple sclerosis: What we have learnt by February 2021 · 2021 · Multiple Sclerosis Journal
Preprints & conceptual frameworks (not peer-reviewed)
- When Local Truth Produces Global Falsehood: Moral Grounding, Institutional Volition, and Systemic Error in Artificial Intelligence · 2026 · Preprints.org
- Measuring What Survives: Pressure-Filtered Persistence and a Masking Index for LLM-Agent Safety Evaluation · 2026 · Research Square
- A Topological Robustness and Anti-Fragility Framework for LLM Agents · 2026 · Research Square
- Human Digital Twins and Self-Knowledge: A Governed Attractor Architecture for Sound Self-Abduction · 2026 · Preprints.org
- From Plausible Narrative to Sound Abduction: A Governed Abductive Architecture for Medical Digital Twins in Multiple Sclerosis Care · 2026 · Preprints.org
- TotalSpineSeg: Robust Spine Segmentation with Landmark-Based Labeling in MRI · 2025 · ResearchGate (preprint) · associated conference material
Public scholarly profiles
- ORCID0009-0004-8538-7140
- Scopus Author ID57223002583
- ResearchGateEmanuel Shirbint
- GitHubEmanuelShirbint
- LinkedInEmanuel Shirbint
- Emailshirbintemanuel@gmail.com