The University of Bergen, in partnership with Adhera Health SL (Spain), invites applications for a PhD Research Fellow position focused on Multimodal Large Language Models for Patient-Generated Data in Shared Decision Making, as part of the ENDOTRAIN Doctoral Network (DC7). This prestigious opportunity is funded by the European Commission through the MSCA Doctoral Network Endotrain and offers a unique interdisciplinary training environment at the intersection of artificial intelligence, digital health, and endocrinology.

The successful candidate will join Europe’s first doctoral network in digital endocrinology, which integrates AI, sensor technology, omics, and clinical medicine to transform the diagnosis and treatment of adrenal diseases. The project is embedded in Work Package 2 – Technologies for Multimodal Data, and aims to advance the development of multimodal Large Language Models (LLMs) that integrate time-series physiological data, clinical assessments, and natural language summaries, with a particular focus on adrenal insufficiency.

Key research tasks include designing and evaluating multimodal AI architectures that combine wearable-derived time-series signals, structured clinical data, and patient-generated language reports; exploring advanced fusion strategies for integrating heterogeneous modalities into context-aware clinical decision support systems; and creating visual and language-based representations of complex patient data to support shared decision making among patients, caregivers, and clinicians. The project involves participatory research with approximately 40 patients living with adrenal insufficiency, utilizing surveys, interviews, and workshops.

The position offers excellent opportunities for international networking and collaboration, including secondments at UiB (Norway), Karolinska Institutet (Sweden), and University Hospital Zurich (Switzerland). The research outputs are expected to contribute to top-tier venues such as AAAI, ACL, NeurIPS, ICML, ACM CHI, and ICHI, and to promote ethical, explainable, and responsible deployment of multimodal LLMs in endocrine disease management.

Applicants must hold a master’s degree (or equivalent) in Mathematics, Computer Science, Computational Linguistics, Engineering, or Medical Engineering, with strong expertise in AI, machine learning, and NLP. Experience with multimodal learning, strong programming skills (Python, PyTorch/TensorFlow), and a demonstrated coding portfolio are required. English proficiency is essential, and compliance with the MSCA mobility rule is mandatory. Desirable qualifications include experience with LLMs, multimodal fusion, explainable AI, digital health, and a record of publications or strong aspiration to publish in leading AI or HCI venues.

The position provides a competitive salary according to MSCA and national regulations, full social security coverage, and a dedicated budget for travel, training, and secondments. Diversity and inclusion are core values of the programme, with active encouragement for women, people with immigrant backgrounds, and people with disabilities to apply.

Applications must be submitted via the Jobbnorge portal, including all mandatory attachments: application form, CV, mobility declaration, and motivation letter. Documentation of the master’s degree is required, and candidates with pending degrees must provide a statement confirming the expected award date. For further information, contact Dr. Luis Fernandez-Luque at Adhera Health SL or refer to the ENDOTRAIN programme webpages.