Postdoctoral Researcher in Local AI/Machine Learning for Offline Archaeology CY Cergy Paris Université in France

CY Cergy Paris Université is advertising a postdoctoral researcher position in local AI / machine learning for offline environments, with a strong application to archaeology and cultural heritage documentation. The post is hosted by ETIS Lab (UMR 8051) in Pontoise, in the Cergy-Pontoise area near Paris, and is part of two closely connected Horizon Europe projects: ECHOES and StratiGraph.

The role sits at the intersection of computer science, machine learning, knowledge graphs, and digital heritage. The research challenge is to develop robust on-device and offline-first AI methods that can work in the field on limited hardware such as laptops, tablets, and embedded devices. The successful candidate will investigate model compression, quantization, pruning, knowledge distillation, few-shot learning, semi/self-supervised learning, domain generalization, domain adaptation, out-of-distribution detection, uncertainty quantification, and intermittent federated learning. A further key direction is knowledge-graph-aware learning, coupling neural models with symbolic constraints and controlled vocabularies to support archaeological documentation workflows.

The projects aim to build tools for validate-as-you-collect field documentation, with real-time AI support for photographs, 3D models, stratigraphic units, and textual records. The work is explicitly oriented toward archaeological field use, including offline processing and later synchronization to the cloud when connectivity returns. Case studies mentioned include Sarmizegetusa Regia in Romania, the Basilica Iulia in Rome, and St. John Lampadistis Monastery in Cyprus.

The appointment is for 12 months, with the possibility of extension, and the salary follows the CY Cergy Paris Université salary scale, commensurate with experience. The position is funded through Horizon Europe projects. The start date is as soon as possible, and the application deadline is 30 August 2026. Applications are reviewed on a rolling basis.

Eligibility requirements include a PhD in Computer Science or a closely related field obtained before the start date, a strong background in one or more relevant ML areas, solid programming and engineering skills, and a publication record in recognized international venues. Knowledge of RDF, SPARQL, and CIDOC-CRM is advantageous, as is experience with constrained hardware or mobile/desktop deployment. Excellent English is required; French is not required. The researcher will collaborate with archaeologists, heritage scientists, and computer scientists across Europe and will be expected to contribute to project deliverables, publish in leading venues, and support PhD and MSc students.