RAISE Doctoral Network for AI-Driven Scientific Discovery

Part B1 - Proposal Template

Call: HORIZON-RAISE-2026-01-03 (submitted via HORIZON-MSCA-2026-DN-01-01)

Deadline: 24 November 2026, 17:00 CET

DOCUMENT LIMIT: Maximum 34 pages (Sections 1-3: max 30 pages starting page 5)


How to use this template

- Replace every [Write your content here] block with your actual proposal text.

- Replace every [PLACEHOLDER] in tables with real data.

- [RAISE GUIDANCE: ...] notes are advisory — remove them before submission.

- All tables are mandatory and must be retained.

- Page budget (informative): TOC ~1 p, Org list ~2 p, Section 1 ~15 p, Section 2 ~8 p, Section 3 ~7 p.

- The body text (from the start of Section 1 through the end of Section 3) must not exceed 30 pages.


TABLE OF CONTENTS

SectionTitlePage
List of Participating OrganisationsConsortium member table; non-academic data table3
1Excellence5
1.1Quality and pertinence of the project's research and innovation objectives5
1.2Soundness of proposed methodology11
1.3Quality and credibility of training programme13
1.4Quality of supervision15
2Impact17
2.1Contribution to structuring doctoral training17
2.2Career perspectives and employability19
2.3Dissemination and exploitation plan20
2.4Magnitude and importance of contribution21
3Quality and Efficiency of Implementation23
3.1Work plan, risks, and effort23
3.2Quality of consortium31

LIST OF PARTICIPATING ORGANISATIONS

Format note: This section counts toward the 2-page limit for organisational information (pages 3–4 of the document). It does NOT count toward the 30-page scientific body.

Consortium Member Table

Legal Entity Short NameAcademic ✓Non-Academic ✓Awards Doctoral Degrees ✓CountryDept / Division / LaboratoryScientist / Person-in-ChargeRole
[INSTITUTION-1][CC][Department / Lab name][Prof. First Last]Coordinator (Beneficiary)
[INSTITUTION-2][CC][Department / Lab name][Prof. First Last]Beneficiary
[INSTITUTION-3][CC][Department / Lab name][Prof. First Last]Beneficiary
[INSTITUTION-4][CC][Department / Lab name][Prof. First Last]Beneficiary
[INSTITUTION-5][CC][Department / Lab name][Prof. First Last]Beneficiary
[INSTITUTION-6][CC][Department / Lab name][Prof. First Last]Beneficiary
[COMPANY-1][CC][Division name][Dr. First Last]Beneficiary
[COMPANY-2][CC][Division name][Dr. First Last]Beneficiary
[INSTITUTE-1][CC][Department / Lab name][Dr. First Last]Associated Partner
[COMPANY-3][CC][Division name][Dr. First Last]Associated Partner
[INSTITUTION-7][CC][Department / Lab name][Prof. First Last]Associated Partner linked to [INSTITUTION-2]

[RAISE GUIDANCE: The consortium should cover a breadth of scientific domains where AI is transformative (e.g., structural biology, climate science, materials discovery, astrophysics, drug design). Include at least 2–3 non-academic beneficiaries (industry, national labs, research infrastructures) to ensure exposure to applied AI-in-science contexts. Ensure geographic spread across EU Member States and Associated Countries. A minimum of 3 beneficiaries from 3 different EU Member States is required; RAISE specifically encourages broad European participation to build the virtual institute vision.**

Data for Non-Academic Beneficiaries

Legal Entity Short NameCountryType of Organisation (SME / Large Enterprise / Public Body / Other)Primary Activity / SectorAnnual Turnover (M€) or Budget (M€)Number of EmployeesMain Role in the Network
[COMPANY-1][CC][Large Enterprise / SME / ...][e.g., Pharmaceutical R&D / AI Software / ...][XX][XXXX][e.g., Industrial secondment host; co-supervision of DC-07, DC-11; provision of proprietary datasets for AI benchmarking]
[COMPANY-2][CC][Large Enterprise / SME / ...][e.g., Scientific Instrumentation / HPC services / ...][XX][XXXX][e.g., Industrial secondment host; co-supervision of DC-09; HPC resource provision for network-wide training]

[RAISE GUIDANCE: Every non-academic beneficiary must demonstrate genuine scientific engagement — not merely administrative or hosting capacity. Describe the specific datasets, compute infrastructure, domain expertise, or translation pathways they bring to AI-in-science research.**


1. EXCELLENCE

(Section 1 target: approximately 15 pages of the 30-page body)

1.1 Quality and pertinence of the project's research and innovation objectives

Introduction, objectives and overview

[Write your content here]

[RAISE GUIDANCE: Open with a compelling framing of the scientific challenge: why does [your chosen scientific domain(s)] require fundamentally new AI capabilities — not incremental applications of existing tools? Position RAISE-[ProjectAcronym] within the broader RAISE (Research and Innovation for AI in Science in Europe) initiative. Articulate how this network contributes to RAISE's vision of a European virtual institute that pools resources, talent, and infrastructure for AI-driven scientific discovery. Reference the strategic relevance to the European Research Area and the competitiveness of European science globally. Recommended length: 0.5–1 page.**

Overall Objective

[Write your content here — one clear sentence stating the overarching scientific and training goal]

Specific Objectives

The specific objectives of [ProjectAcronym] are:

#Specific ObjectiveAddressed in WP(s)
SO1[Write your content here — e.g., Develop novel deep learning architectures for [scientific problem], surpassing the state of the art in [measurable metric]]WP[X]
SO2[Write your content here — e.g., Create open benchmark datasets and evaluation frameworks for AI in [domain]]WP[X]
SO3[Write your content here — e.g., Train 15 doctoral candidates at the frontier of AI and [scientific domain]]WP[X], WP[X]
SO4[Write your content here — e.g., Validate AI-developed tools in real-world scientific workflows at [type of facility/partner]]WP[X]
SO5[Write your content here — e.g., Build a sustainable European network for AI in [domain] research and training]WP[X]

[RAISE GUIDANCE: Objectives must be SMART (Specific, Measurable, Achievable, Relevant, Time-bound). Each objective should be traceable to one or more DCs and to deliverables/milestones. Avoid vague language such as "improve" or "explore" without quantifiable targets.**

State of the Art and Gaps Addressed

[Write your content here]

[RAISE GUIDANCE: Provide a rigorous critical review of the state of the art in (a) the scientific domain(s) and (b) the AI/ML methodologies. Identify the specific gaps that current AI tools cannot address and explain WHY closing these gaps requires the development of new AI systems — not re-application of existing ones. This is the core intellectual justification for the RAISE eligibility criterion: AI must be integral and innovative, not merely instrumental.**

Beyond the State of the Art

[Write your content here]

[RAISE GUIDANCE: Explicitly describe what AI innovations each DC project will produce. Use language such as: "DC-03 will develop a novel physics-informed neural network architecture that..." rather than "DC-03 will use machine learning to analyse...". Reviewers will scrutinise whether AI development is genuinely new.**


Individual DC Research Projects

[RAISE GUIDANCE: Each DC project must demonstrate that AI development is INDISPENSABLE to the scientific research goal — the DC must be developing innovative AI systems, models, algorithms, or tools for the scientific domain, not merely applying existing AI methods as a secondary technique. The scientific and AI components must be inseparable. Ensure secondments are scientifically justified and that the non-academic secondment (if applicable) provides genuine AI-in-science exposure.**

DC Project Summary Table

DC #Host InstitutionProject TitlePrimary Scientific DomainAI Innovation ElementExpected ResultsSecondment PartnersDuration (months)
DC-01[INSTITUTION-X][Write your content here][e.g., Structural Biology][e.g., Equivariant graph neural networks for protein-ligand docking][Write your content here][INSTITUTION-Y] (M13–18); [COMPANY-Z] (M19–24)36
DC-02[INSTITUTION-X][Write your content here][e.g., Climate Science][e.g., Physics-informed transformer models for sub-grid parameterisation][Write your content here][INSTITUTION-Y] (M7–12); [INSTITUTION-Z] (M25–30)36
DC-03[INSTITUTION-X][Write your content here][e.g., Materials Discovery][e.g., Generative diffusion models for inverse materials design][Write your content here][COMPANY-Y] (M13–18)36
DC-04[INSTITUTION-X][Write your content here][e.g., Astrophysics][e.g., Simulation-based inference for gravitational wave parameter estimation][Write your content here][INSTITUTION-Y] (M19–24); [INSTITUTE-Z] (M7–12)36
DC-05[INSTITUTION-X][Write your content here][e.g., Drug Discovery][e.g., Foundation model fine-tuning for ADMET property prediction][Write your content here][COMPANY-Y] (M13–24)36
DC-06[INSTITUTION-X][Write your content here][e.g., Genomics][e.g., Attention-based multi-omics integration for regulatory element prediction][Write your content here][INSTITUTION-Y] (M7–12); [COMPANY-Z] (M25–30)36
DC-07[INSTITUTION-X][Write your content here][e.g., High-Energy Physics][e.g., Fast surrogate models for Monte Carlo event generation][Write your content here][INSTITUTE-Y] (M13–18); [COMPANY-Z] (M19–24)36
DC-08[INSTITUTION-X][Write your content here][e.g., Neuroscience][e.g., Latent variable models for large-scale neural population dynamics][Write your content here][INSTITUTION-Y] (M7–18)36
DC-09[INSTITUTION-X][Write your content here][e.g., Earth Observation][e.g., Self-supervised learning for multi-modal satellite data fusion][Write your content here][COMPANY-Y] (M13–24); [INSTITUTION-Z] (M7–12)36
DC-10[INSTITUTION-X][Write your content here][e.g., Chemistry][e.g., Reinforcement learning for autonomous reaction optimisation][Write your content here][COMPANY-Y] (M19–24); [INSTITUTION-Z] (M13–18)36
DC-11[INSTITUTION-X][Write your content here][e.g., Medical Imaging][e.g., Uncertainty-aware convolutional architectures for pathology segmentation][Write your content here][COMPANY-Y] (M13–18); [INSTITUTION-Z] (M25–30)36
DC-12[INSTITUTION-X][Write your content here][e.g., Ecology][e.g., Neural ODEs for ecosystem dynamics modelling from sparse sensor data][Write your content here][INSTITUTION-Y] (M7–12); [COMPANY-Z] (M19–24)36
DC-13[INSTITUTION-X][Write your content here][e.g., Fluid Dynamics][e.g., Neural operators for turbulence closure in high-Reynolds-number flows][Write your content here][INSTITUTION-Y] (M13–18); [INSTITUTE-Z] (M25–30)36
DC-14[INSTITUTION-X][Write your content here][e.g., Bioinformatics][e.g., Geometric deep learning for RNA tertiary structure prediction][Write your content here][INSTITUTION-Y] (M7–12); [COMPANY-Z] (M13–18)36
DC-15[INSTITUTION-X][Write your content here][e.g., Quantum Science][e.g., Machine learning quantum error mitigation for near-term quantum devices][Write your content here][INSTITUTION-Y] (M19–24); [COMPANY-Z] (M25–30)36

Note on secondments: Each DC should have at least one planned secondment. Secondments at non-academic partners must be included for DCs where knowledge transfer to industry/society is scientifically relevant. Total secondment duration per DC should not exceed 30% of the DC's contract duration (i.e., max ~11 months for a 36-month DC).

Detailed Individual DC Descriptions

[RAISE GUIDANCE: For each DC, write a self-contained ~0.5-page description covering: (1) scientific background and gap, (2) AI innovation — what new AI system/method will be developed and why it does not exist yet, (3) research objectives and methodology, (4) expected results and their scientific/technological significance, (5) secondment plan and its scientific rationale. Total for all 15 DC descriptions: approximately 6–7 pages.**

DC-01: [Title]

[Write your content here]

DC-02: [Title]

[Write your content here]

DC-03: [Title]

[Write your content here]

DC-04: [Title]

[Write your content here]

DC-05: [Title]

[Write your content here]

DC-06: [Title]

[Write your content here]

DC-07: [Title]

[Write your content here]

DC-08: [Title]

[Write your content here]

DC-09: [Title]

[Write your content here]

DC-10: [Title]

[Write your content here]

DC-11: [Title]

[Write your content here]

DC-12: [Title]

[Write your content here]

DC-13: [Title]

[Write your content here]

DC-14: [Title]

[Write your content here]

DC-15: [Title]

[Write your content here]


Pertinence and Innovative Aspects

[Write your content here]

[RAISE GUIDANCE: Explain the collective innovation beyond what any single DC project achieves alone. Describe how the portfolio of 15 DCs creates synergies: shared datasets, shared AI model components, cross-domain benchmarking, and joint tools that benefit the entire RAISE initiative. Reference alignment with European AI strategy, the European Research Area priorities, and EU missions where relevant.**


1.2 Soundness of proposed methodology

Overall Methodology

[Write your content here]

[RAISE GUIDANCE: Describe the network's unified methodological framework for AI-in-science research. This should address: (a) how AI development methodology (e.g., ML engineering pipeline, model validation, reproducibility standards) is integrated with scientific domain methodology; (b) how the network ensures scientific rigour in AI claims (avoiding over-optimistic benchmark results, reporting negative results, using proper baselines); (c) how the network handles data — collection, curation, annotation, quality control, and sharing. Cite relevant methodological standards (e.g., NeurIPS reproducibility checklist, domain-specific reporting guidelines).**

Integration of Methods and Disciplines

[Write your content here]

[RAISE GUIDANCE: Demonstrate genuine interdisciplinarity: each DC sits at the intersection of at least one scientific domain and AI/ML. Describe how the training programme and joint research activities will prevent DCs from becoming isolated specialists in only one domain. Explain the specific mechanisms — joint seminars, cross-DC collaboration tasks, shared reading groups — that build integrative competence.**

Gender Dimension and Other Diversity Aspects

[Write your content here]

[RAISE GUIDANCE: MSCA requires reflection on how gender/sex and other diversity dimensions are incorporated into the research content (not just HR practices). For AI in science: address algorithmic bias in AI models trained on historically non-representative scientific datasets; data diversity and representativeness; fairness in AI-assisted scientific discovery (e.g., AI models trained predominantly on data from high-income countries). Also include the network's own diversity targets for DC recruitment (minimum 40% of underrepresented gender). Mention planned unconscious bias training for supervisors.**

Open Science Practices

[Write your content here]

[RAISE GUIDANCE: RAISE networks are expected to set a high standard for open science. Specify: (1) Open access publication commitment (Plan S / diamond open access); (2) FAIR data management — link to the Data Management Plan in Part B2; (3) Open-source release policy for AI models, training code, and benchmark datasets developed by DCs; (4) Pre-registration of computational experiments where applicable; (5) Use of open science platforms (e.g., Zenodo, HuggingFace, OpenML, Papers with Code). The release of open AI tools for the scientific community is a key RAISE deliverable.**


1.3 Quality and credibility of training programme

Overview and Content Structure

[Write your content here]

[RAISE GUIDANCE: The training programme must be structured at doctoral level and must include DEDICATED training on AI in science — not generic programming or data science. This means: advanced ML theory, AI model development methodology, evaluation and benchmarking of AI systems in scientific contexts, responsible AI in science, and AI reproducibility. Distinguish between (a) Network-Wide Training Events (NWTEs, attended by all DCs), (b) local/institutional training at the host, and (c) transferable skills training. The total training load should be realistic — typically 15–30 ECTS over the doctoral period. Map training to career development stages.**

Training Programme Structure

Training ModuleTypeTarget DCsECTSDelivery ModeMonth(s)
AI Foundations for Science (methods, architectures, theory)NWTEAll3Residential schoolM4–5
Scientific Computing and HPC for AINWTEAll2Residential schoolM8
Research Methods and Scientific WritingNWTEAll2WorkshopM6
Responsible AI in Science (bias, fairness, ethics, reproducibility)NWTEAll2WorkshopM12
Advanced AI Seminar Series (invited speakers, DC presentations)NWTEAll2Virtual + residentialM1–36
Industry Immersion WorkshopNWTEAll2Residential at industry partnerM18
Entrepreneurship and InnovationNWTEAll1WorkshopM24
Science Communication and PolicyNWTEAll1WorkshopM30
Domain-specific doctoral training (local)LocalBy host3Institutional coursesM1–12
Transferable skills (local)LocalBy host2Institutional coursesM1–36
Total20

Table 1: Main Network-Wide Training Events

Event #TitleTypeECTSLead InstitutionParticipating InstitutionsAction MonthLocationDuration (days)
NWTE-1Kick-off Meeting and Network InductionKick-off + induction0.5[INSTITUTION-1]AllM1[City, Country]3
NWTE-2Summer School I: AI Foundations for Scientific DiscoveryDoctoral school3[INSTITUTION-2]AllM5[City, Country]5
NWTE-3Workshop: Scientific Computing and HPC for AITechnical workshop2[INSTITUTION-3]AllM8[City, Country]3
NWTE-4Workshop: Research Methods, Open Science, and Scientific WritingSkills workshop2[INSTITUTION-4]AllM10[City, Country]3
NWTE-5Annual Network Meeting I + DC Progress SymposiumScientific meeting1[INSTITUTION-5]AllM12[City, Country]2
NWTE-6Workshop: Responsible AI in Science — Ethics, Bias, and ReproducibilitySkills + methods2[INSTITUTION-1]AllM14[City, Country]2
NWTE-7Summer School II: Advanced Topics in AI for ScienceAdvanced doctoral school3[INSTITUTION-3]AllM18[City, Country]5
NWTE-8Industry Immersion Week at [COMPANY-1 / COMPANY-2]Industry engagement2[COMPANY-1]AllM20[City, Country]5
NWTE-9Annual Network Meeting II + Midterm Scientific SymposiumScientific meeting1[INSTITUTION-6]AllM24[City, Country]2
NWTE-10Workshop: Entrepreneurship, Technology Transfer, and Spin-outTransferable skills1[COMPANY-2]AllM26[City, Country]2
NWTE-11Science Communication, Outreach, and Policy EngagementTransferable skills1[INSTITUTION-2]AllM30[City, Country]2
NWTE-12Final Network Conference: AI in Science — Results and OutlookDissemination conference1[INSTITUTION-1]All + invited guestsM35[City, Country]3

[RAISE GUIDANCE: NWTE-2 and NWTE-7 (the summer schools) are the flagship training events and should be described in detail in the body text — covering content, invited instructors, and how they build AI-in-science competence. The Industry Immersion Week (NWTE-8) should be described with specific activities at the company partner (e.g., working on real industrial AI-in-science problems). NWTE-6 on Responsible AI is specifically encouraged by RAISE to ensure DCs are equipped to address societal and ethical dimensions of their work.**

Individual Development Plans (IDPs)

[Write your content here]

[RAISE GUIDANCE: Describe the IDP process: initial skills assessment at M1, annual review, supervisor and DC co-ownership of the IDP. The IDP should explicitly track AI-in-science competence development alongside domain expertise. Mention the network's commitment to the European Charter for Researchers and Code of Conduct for the Recruitment of Researchers.**


1.4 Quality of supervision

Qualifications and Experience

[Write your content here]

[RAISE GUIDANCE: For each principal supervisor (and co-supervisor where relevant), briefly describe: (a) track record in AI research OR in the scientific domain (at least one supervisor per DC must have strong AI/ML expertise); (b) previous doctoral supervision record (number of completed PhDs); (c) experience with international, interdisciplinary projects. It is critical that each DC has at least one supervisor with genuine AI development expertise — not merely AI application experience. Provide a table of supervisors.**

Table: Supervisory Team Overview

DC #Principal SupervisorInstitutionAI ExpertiseDomain ExpertiseNo. of Completed PhD SupervisionsKey Publications (AI or Domain, last 5 years)
DC-01[Prof. First Last][INSTITUTION-X][e.g., Graph neural networks, geometric deep learning][e.g., Structural biology, protein folding][X][Write your content here]
DC-02[Prof. First Last][INSTITUTION-X][e.g., Physics-informed ML, scientific machine learning][e.g., Climate modelling, NWP][X][Write your content here]
DC-03[Prof. First Last][INSTITUTION-X][e.g., Generative models, diffusion models][e.g., Computational materials science][X][Write your content here]
DC-04[Prof. First Last][INSTITUTION-X][e.g., Simulation-based inference, normalising flows][e.g., Gravitational wave astronomy][X][Write your content here]
DC-05[Prof. First Last][INSTITUTION-X][e.g., Foundation models, transfer learning][e.g., Computational drug discovery][X][Write your content here]
DC-06[Prof. First Last][INSTITUTION-X][e.g., Attention mechanisms, multi-modal learning][e.g., Genomics, regulatory genomics][X][Write your content here]
DC-07[Prof. First Last][INSTITUTION-X][e.g., Surrogate modelling, emulation][e.g., High-energy physics, detector simulation][X][Write your content here]
DC-08[Prof. First Last][INSTITUTION-X][e.g., Latent variable models, variational inference][e.g., Computational neuroscience][X][Write your content here]
DC-09[Prof. First Last][INSTITUTION-X][e.g., Self-supervised learning, contrastive learning][e.g., Remote sensing, Earth observation][X][Write your content here]
DC-10[Prof. First Last][INSTITUTION-X][e.g., Reinforcement learning, Bayesian optimisation][e.g., Synthetic chemistry, reaction design][X][Write your content here]
DC-11[Prof. First Last][INSTITUTION-X][e.g., Uncertainty quantification, Bayesian deep learning][e.g., Medical image analysis, pathology][X][Write your content here]
DC-12[Prof. First Last][INSTITUTION-X][e.g., Neural ODEs, continuous-time models][e.g., Ecology, biodiversity informatics][X][Write your content here]
DC-13[Prof. First Last][INSTITUTION-X][e.g., Neural operators, operator learning][e.g., Computational fluid dynamics, turbulence][X][Write your content here]
DC-14[Prof. First Last][INSTITUTION-X][e.g., Geometric deep learning, equivariant networks][e.g., RNA bioinformatics, structural bioinformatics][X][Write your content here]
DC-15[Prof. First Last][INSTITUTION-X][e.g., Quantum ML, variational quantum algorithms][e.g., Quantum information, quantum error correction][X][Write your content here]

Quality of Supervision Arrangements

[Write your content here]

[RAISE GUIDANCE: Describe the formal supervision structure: thesis advisory committee composition (including at least one external member and, where possible, one non-academic member), meeting frequency, escalation pathway for supervisory conflicts, and the network's monitoring role via the Training and Supervision Committee. Explain how co-supervision across institutions is organised — joint supervision agreements, cost sharing, and IP arrangements. Align with the MSCA Charter and Code commitments.**


2. IMPACT

(Section 2 target: approximately 8 pages of the 30-page body)

2.1 Contribution to structuring doctoral training

Non-Academic Sector Contribution

[Write your content here]

[RAISE GUIDANCE: Describe how non-academic partners (industry, national labs, research infrastructures) contribute substantively to doctoral training — not merely as secondment hosts. This includes: co-designing training modules (especially NWTE-8 and domain-specific sessions), co-supervising DCs, providing access to proprietary datasets or computing facilities, contributing to the teaching programme of summer schools, and offering mentoring beyond the formal secondment period. For RAISE, the non-academic dimension should emphasise the AI translation pipeline: how industry partners help DCs understand the path from AI research prototype to deployed scientific tool.**

Sustainable Elements After Funding Ends

[Write your content here]

[RAISE GUIDANCE: RAISE networks are expected to contribute to the long-term RAISE virtual institute infrastructure. Describe concrete sustainability plans: (a) continuation of the annual network conference as an open community event; (b) open-source maintenance plan for AI tools and datasets created by the network; (c) institutional commitments to continue PhD training in AI-in-science after the grant (e.g., new joint PhD programmes, chairs, centres); (d) plans to integrate network-developed training materials into permanent graduate curricula; (e) engagement with the RAISE data and compute infrastructure for continued use by the scientific community.**


2.2 Career perspectives and employability

Impact on Fellows' Careers

[Write your content here]

[RAISE GUIDANCE: The career landscape for PhD graduates with combined AI and domain science expertise is exceptionally strong. Describe the range of career pathways: (a) academic research (AI-focused or domain-science-focused, or bridging both); (b) industrial R&D in AI-for-science companies, pharma, tech; (c) national laboratories, research infrastructures, and large-scale scientific facilities; (d) policy, technology transfer, and start-up creation. Describe how the network's training, mentoring, and industrial exposure specifically prepare DCs for these pathways. Include plans for career tracking of DCs after the network ends (5-year follow-up survey). Mention the alumninetwork/community-building activities.**


2.3 Dissemination and exploitation plan

Plan for Dissemination, Exploitation, and Communication

[Write your content here]

[RAISE GUIDANCE: Structure the plan around three distinct activities: (1) DISSEMINATION — targeting the scientific community: open-access publications, conference presentations, preprints, AI model releases on HuggingFace/Zenodo/GitHub; (2) EXPLOITATION — targeting economic/technological uptake: engagement with industrial partners beyond the consortium, technology licensing, spin-out creation; (3) COMMUNICATION — targeting the general public, policymakers, and students: press releases, social media, public lectures, science fairs, policy briefs. Define KPIs for each activity (e.g., number of open-access papers, GitHub stars on network repositories, public engagement events per year).**

Key Performance Indicators for Dissemination and Communication

ActivityKPITarget (over 4 years)
Open-access journal publicationsNumber of peer-reviewed papers≥ 45 (≥ 3 per DC)
High-impact venue publications (NeurIPS, ICML, Nature family, Science)Number of papers≥ 10
Open-source AI tools/models releasedNumber of GitHub repositories with active maintenance≥ 15
Open datasets published on Zenodo/other FAIR repositoryNumber of datasets≥ 10
Conference presentations (oral + poster)Number≥ 60
Network website unique visitorsAnnual≥ 5,000
Social media reach (LinkedIn, Twitter/X, Mastodon)Followers≥ 2,000
Public outreach eventsNumber≥ 20
Policy briefs or contributions to consultationsNumber≥ 3
Media coverage (press, online news)Number of items≥ 15

IP Management Strategy

[Write your content here]

[RAISE GUIDANCE: Address IP proactively. Key points: (a) By default, IP generated by DCs during secondments must be clearly allocated — specify in the Consortium Agreement and individual secondment agreements; (b) For AI models trained on proprietary industrial datasets, address ownership and licensing terms explicitly; (c) Default to open-source licensing (MIT/Apache 2.0) for AI tools developed using only public data and public funding; (d) When industrial partners co-develop tools, describe the licensing model that still enables academic publication; (e) Describe the role of the institutional technology transfer offices.**


2.4 Magnitude and importance of contribution

Expected Scientific Impact(s)

[Write your content here]

[RAISE GUIDANCE: Articulate the scientific impact at three levels: (1) WITHIN each scientific domain — what specific discoveries, capabilities, or paradigm shifts will the AI tools developed enable that were not possible before? Give concrete examples tied to specific DC projects. (2) FOR AI research — what novel AI methods, architectures, or theoretical insights will the network produce that advance the field of AI itself? (3) FOR THE RAISE INITIATIVE — how does this network contribute to building the RAISE virtual institute's scientific portfolio, shared infrastructure, and pan-European AI-in-science ecosystem? Emphasise that the network's outputs will be freely available to the entire European scientific community, amplifying impact beyond the consortium.**

Expected Economic and Technological Impact(s)

[Write your content here]

[RAISE GUIDANCE: Connect AI-in-science research to economic value. Describe: (a) How AI tools developed in the network will reduce costs or accelerate timelines in specific industrial sectors (drug discovery, materials development, climate services, etc.); (b) Spin-out and start-up potential — are any DC projects in areas with commercial applications? (c) Contribution to European technological sovereignty in AI for science — reducing dependence on non-European AI platforms for critical scientific applications; (d) HPC and AI infrastructure utilisation — does the network contribute to valorising European compute infrastructure investments (e.g., EuroHPC)?**

Expected Societal Impact(s)

[Write your content here]

[RAISE GUIDANCE: The societal impact of AI-in-science research can be profound but must be articulated carefully and specifically. Address: (a) How AI-accelerated scientific discovery in the network's domains will contribute to societal challenges (health, climate, energy, food security — as applicable to your specific DC portfolio); (b) Equity and access dimensions: open-source tools make AI-driven science accessible to researchers in lower-resourced institutions; (c) Responsible AI practices embedded in the training programme ensure DCs will be ethical practitioners; (d) Public engagement with AI and science — countering AI hype and building scientifically literate public understanding of AI capabilities and limitations; (e) Contribution to training the next generation of responsible AI scientists for Europe.**


3. QUALITY AND EFFICIENCY OF IMPLEMENTATION

(Section 3 target: approximately 7 pages of the 30-page body)

3.1 Work plan, risks, and effort

[Write your content here — brief introduction to the work plan structure]

[RAISE GUIDANCE: The work plan should have a logical architecture: typically WP1 = Management and Coordination; WP2–WPn = Scientific Research WPs (one per domain cluster or AI methodology cluster); WPn+1 = Training and Supervision; WPn+2 = Dissemination, Exploitation, and Communication. Each DC must be unambiguously assigned to one primary scientific WP. Ensure the Gantt chart (if included as a figure) clearly shows DC start months, NWTE timing, deliverable and milestone due dates.**


Table 3.1a: Work Packages

WP #TitleStart MonthEnd MonthLead ParticipantParticipating InstitutionsDCs InvolvedObjectivesDescription
WP1Project Management and Coordination148[INSTITUTION-1]AllAllO1: Ensure efficient day-to-day management; O2: Ensure financial and ethical compliance; O3: Coordinate across WPs and between DCs[Write your content here — describe governance structure, management bodies (Steering Committee, Training and Supervision Committee, Ethics Board), reporting and monitoring procedures, internal communication tools, and contingency planning]
WP2[AI Research Domain Cluster 1, e.g., AI for Molecular Science]148[INSTITUTION-2][INSTITUTION-X, INSTITUTION-Y, COMPANY-Z]DC-01, DC-05, DC-06, DC-14O4: Develop AI models for [domain cluster]; O5: Produce open benchmarks for [domain cluster][Write your content here — describe the unifying scientific questions, shared methodological infrastructure (datasets, compute, code), and how the DCs in this cluster collaborate and complement each other]
WP3[AI Research Domain Cluster 2, e.g., AI for Earth and Environmental Science]148[INSTITUTION-3][INSTITUTION-X, INSTITUTION-Y, COMPANY-Z]DC-02, DC-09, DC-12, DC-13O6: [Write your content here]; O7: [Write your content here][Write your content here]
WP4[AI Research Domain Cluster 3, e.g., AI for Physics and Astronomy]148[INSTITUTION-4][INSTITUTION-X, INSTITUTION-Y, INSTITUTE-Z]DC-04, DC-07, DC-15O8: [Write your content here]; O9: [Write your content here][Write your content here]
WP5[AI Research Domain Cluster 4, e.g., AI for Biomedical Science]148[INSTITUTION-5][INSTITUTION-X, COMPANY-Y, COMPANY-Z]DC-03, DC-08, DC-10, DC-11O10: [Write your content here]; O11: [Write your content here][Write your content here]
WP6Training, Supervision, and Career Development148[INSTITUTION-6]AllAllO12: Deliver all NWTEs; O13: Coordinate IDPs; O14: Monitor DC progress and supervisory quality[Write your content here — describe NWTE planning and coordination procedures, IDP framework, supervisory monitoring, the role of thesis advisory committees, and the training quality assurance process]
WP7Dissemination, Exploitation, and Communication148[INSTITUTION-1]AllAllO15: Maximise scientific dissemination; O16: Drive exploitation of results; O17: Communicate with public and policymakers[Write your content here — describe the dissemination strategy for publications and open software, the exploitation strategy for industrial uptake, and the communication strategy for public engagement]

Table 3.1b: Deliverables List

Scientific Deliverables

Del. #TitleDescriptionWPLead ParticipantTypeDissemination LevelDue Date (month)
D2.1Open Dataset I: [Domain Cluster 1 benchmark dataset name][Write your content here — describe the dataset: domain, size, format, and significance for AI benchmarking in this domain]WP2[INSTITUTION-2]Dataset (FAIR)PublicM18
D2.2Open-Source AI Toolkit I: [Toolkit name for Domain Cluster 1][Write your content here — describe the toolkit: algorithms, models, documentation, and intended user community]WP2[INSTITUTION-2]Software (open source)PublicM36
D2.3Scientific Report: AI methods for [Domain Cluster 1] — state of the art and network advances[Write your content here]WP2[INSTITUTION-2]Report / white paperPublicM42
D3.1Open Dataset II: [Domain Cluster 2 benchmark dataset name][Write your content here]WP3[INSTITUTION-3]Dataset (FAIR)PublicM20
D3.2Open-Source AI Toolkit II: [Toolkit name for Domain Cluster 2][Write your content here]WP3[INSTITUTION-3]Software (open source)PublicM38
D3.3Scientific Report: AI methods for [Domain Cluster 2] — network advances[Write your content here]WP3[INSTITUTION-3]Report / white paperPublicM44
D4.1Open Dataset III: [Domain Cluster 3 benchmark dataset name][Write your content here]WP4[INSTITUTION-4]Dataset (FAIR)PublicM20
D4.2Open-Source AI Toolkit III: [Toolkit name for Domain Cluster 3][Write your content here]WP4[INSTITUTION-4]Software (open source)PublicM36
D4.3Scientific Report: AI methods for [Domain Cluster 3] — network advances[Write your content here]WP4[INSTITUTION-4]Report / white paperPublicM44
D5.1Open Dataset IV: [Domain Cluster 4 benchmark dataset name][Write your content here]WP5[INSTITUTION-5]Dataset (FAIR)PublicM18
D5.2Open-Source AI Toolkit IV: [Toolkit name for Domain Cluster 4][Write your content here]WP5[INSTITUTION-5]Software (open source)PublicM38
D5.3Scientific Report: AI methods for [Domain Cluster 4] — network advances[Write your content here]WP5[INSTITUTION-5]Report / white paperPublicM44

Management Deliverables

Del. #TitleDescriptionWPLead ParticipantTypeDissemination LevelDue Date (month)
D1.1Data Management Plan (DMP) v1Initial DMP covering all anticipated research data types, FAIR compliance strategy, data sharing and archiving planWP1[INSTITUTION-1]PlanRestrictedM3
D1.2Consortium AgreementSigned CA covering IP, liability, DC employment conditions, secondment agreements, and governanceWP1[INSTITUTION-1]AgreementRestrictedM3
D1.3Data Management Plan (DMP) v2 (updated)Updated DMP reflecting actual data generated at midtermWP1[INSTITUTION-1]PlanRestrictedM24
D1.4Annual Management Report Year 1Progress, financial, and risk management report for Year 1WP1[INSTITUTION-1]ReportRestrictedM12
D1.5Annual Management Report Year 2Progress, financial, and risk management report for Year 2WP1[INSTITUTION-1]ReportRestrictedM24
D1.6Annual Management Report Year 3Progress, financial, and risk management report for Year 3WP1[INSTITUTION-1]ReportRestrictedM36
D1.7Final Management and Coordination ReportComprehensive final report covering all management activities, lessons learned, and sustainability planWP1[INSTITUTION-1]ReportPublicM48
D6.1Training Programme PlanDetailed schedule and content of all NWTEs, IDP framework documentation, and supervisory guidelinesWP6[INSTITUTION-6]PlanPublicM3
D6.2Training Quality Report (Midterm)Assessment of training quality based on DC feedback and supervisory evaluation at midtermWP6[INSTITUTION-6]ReportRestrictedM24
D6.3Training Quality Report (Final)Final assessment of training programme quality and recommendations for future networksWP6[INSTITUTION-6]ReportPublicM46
D7.1Communication and Dissemination StrategyPlan covering target audiences, channels, KPIs, and timeline for all dissemination and communication activitiesWP7[INSTITUTION-1]PlanPublicM3
D7.2Network Website and Social Media ChannelsLaunch of [ProjectAcronym] public website and social media presenceWP7[INSTITUTION-1]Website / onlinePublicM4
D7.3Midterm Dissemination and Exploitation ReportInventory of publications, software releases, datasets, patents, and communication activities at midtermWP7[INSTITUTION-1]ReportPublicM24
D7.4Final Dissemination, Exploitation, and Communication ReportComprehensive final report on all DECs activities; open software/data inventory; exploitation and sustainability planWP7[INSTITUTION-1]ReportPublicM47

Table 3.1c: Milestones List

MS #TitleRelated WP(s)Lead ParticipantDue Date (month)Means of Verification
MS1Project start: all DCs recruited and employedWP1, WP6[INSTITUTION-1]M6All DC employment contracts signed and active; confirmation from beneficiary HR offices
MS2Consortium Agreement signed by all beneficiariesWP1[INSTITUTION-1]M3Signed CA submitted to EC Project Officer
MS3Data Management Plan v1 approvedWP1[INSTITUTION-1]M4DMP approved by Steering Committee; submitted as D1.1
MS4All IDPs established for all 15 DCsWP6[INSTITUTION-6]M6IDPs signed by DC and supervisory team; deposited in network IDP system
MS5NWTE-2 (Summer School I) completedWP6[INSTITUTION-2]M5Attendance records; DC feedback report submitted to Training Committee
MS6First DC publications submitted (preprints)WP2–WP5AllM12Minimum 5 preprints posted on arXiv or equivalent open repository
MS7NWTE-5 (Annual Meeting I) completedWP6[INSTITUTION-5]M12Meeting report and DC presentation abstracts available on project website
MS8All secondment agreements signedWP1, WP6[INSTITUTION-1]M9Signed secondment agreements for all planned DC secondments on file
MS9Open Dataset I and II released (FAIR)WP2, WP3[INSTITUTION-2], [INSTITUTION-3]M20Datasets published on Zenodo with DOI and FAIR metadata; announced on project website
MS10Midterm scientific review by external advisory boardWP1[INSTITUTION-1]M24External Advisory Board report received; recommendations addressed by Steering Committee
MS11NWTE-9 (Annual Meeting II + Midterm Symposium) completedWP6[INSTITUTION-6]M24Meeting report; DC mid-term assessment forms completed
MS12All four open-source AI toolkits in beta releaseWP2–WP5Domain WP leadsM30Beta releases available on GitHub with documentation; announced on project website
MS13Open Datasets III and IV released (FAIR)WP4, WP5[INSTITUTION-4], [INSTITUTION-5]M22Datasets published on Zenodo with DOI; announced on project website
MS14≥ 30 peer-reviewed open-access publications submittedWP7[INSTITUTION-1]M36Publications list verified against open-access repositories
MS15All DC thesis defences completed or on scheduleWP6[INSTITUTION-6]M42Written confirmation from PhD-awarding institutions of defence scheduling or completion
MS16Final Network Conference heldWP7[INSTITUTION-1]M35Conference programme, attendance list, and proceedings/abstracts book available online
MS17All four open-source AI toolkits in stable releaseWP2–WP5Domain WP leadsM40Stable releases (v1.0+) on GitHub; archived on Zenodo with DOI; usage documentation complete
MS18Project completion: all deliverables submittedWP1[INSTITUTION-1]M48All deliverables confirmed received by EC; final financial statement submitted

Table 3.1d: DC Table

DC #Recruiting ParticipantPhD-Awarding EntityStart MonthDuration (months)Secondment Duration (months total)Non-Academic Secondment Duration (months)Scientific DomainAI Innovation Theme
DC-01[INSTITUTION-X][INSTITUTION-X]M136126[e.g., Structural Biology][e.g., Equivariant GNNs]
DC-02[INSTITUTION-X][INSTITUTION-X]M136120[e.g., Climate Science][e.g., Physics-informed transformers]
DC-03[INSTITUTION-X][INSTITUTION-X]M23666[e.g., Materials Discovery][e.g., Generative diffusion models]
DC-04[INSTITUTION-X][INSTITUTION-X]M136120[e.g., Astrophysics][e.g., Simulation-based inference]
DC-05[INSTITUTION-X][INSTITUTION-X]M2361212[e.g., Drug Discovery][e.g., Foundation model fine-tuning]
DC-06[INSTITUTION-X][INSTITUTION-X]M136126[e.g., Genomics][e.g., Multi-modal attention]
DC-07[INSTITUTION-X][INSTITUTION-X]M236126[e.g., High-Energy Physics][e.g., Fast surrogate models]
DC-08[INSTITUTION-X][INSTITUTION-X]M136120[e.g., Neuroscience][e.g., Latent variable models]
DC-09[INSTITUTION-X][INSTITUTION-X]M2361212[e.g., Earth Observation][e.g., Self-supervised multi-modal learning]
DC-10[INSTITUTION-X][INSTITUTION-X]M136126[e.g., Chemistry][e.g., Reinforcement learning for synthesis]
DC-11[INSTITUTION-X][INSTITUTION-X]M23666[e.g., Medical Imaging][e.g., Uncertainty-aware deep learning]
DC-12[INSTITUTION-X][INSTITUTION-X]M136126[e.g., Ecology][e.g., Neural ODEs for dynamics]
DC-13[INSTITUTION-X][INSTITUTION-X]M236120[e.g., Fluid Dynamics][e.g., Neural operators]
DC-14[INSTITUTION-X][INSTITUTION-X]M13666[e.g., Bioinformatics][e.g., Geometric deep learning for RNA]
DC-15[INSTITUTION-X][INSTITUTION-X]M236126[e.g., Quantum Science][e.g., ML for quantum error mitigation]

Notes on Table 3.1d:

- Duration must be between 3 and 36 months.

- Total secondment duration per DC should not exceed 30% of the DC's contract duration (≤ 11 months for a 36-month DC).

- The non-academic secondment duration is a subset of the total secondment duration.

- If a DC does not have a non-academic secondment, enter 0.

- The PhD-awarding entity must be a beneficiary that awards doctoral degrees (as marked in the Consortium Member Table above).


Table 3.1e: Project Risks

Risk #Risk DescriptionLikelihood (Low / Medium / High)Severity (Low / Medium / High)Affected WP(s)Mitigation Strategy
R1DC recruitment failure: qualified candidates with combined AI + domain expertise are scarceMediumHighWP1, WP6, All scientific WPsEarly launch of recruitment (M-3 before project start); active targeting of MSc graduates from AI and domain science programmes; use of MSCA EURAXESS portal and domain-specific job boards; flexible admission criteria recognising AI or domain background with training to bridge the gap; fallback recruitment from associated partner networks
R2DC dropout or prolonged sick leaveLow–MediumHighWP1, WP6, Affected scientific WPIDP monitoring and early intervention; close supervisor support; contingency budget for recruitment of replacement DC (within MSCA rules); clear procedures for extending DC contract within grant duration
R3Supervisor departure or capacity reductionLowHighWP6, Affected scientific WPCo-supervision model ensures at least 2 supervisors per DC; successor supervisor identified within beneficiary institution; Steering Committee oversight of supervisory quality
R4AI research results fall short of objectives (negative or null results)MediumMediumScientific WPsBuilt-in interim assessment milestones (MS10 at M24); early pivoting of research direction allowed within scientific WPs; RAISE networks should normalise and publish negative results as part of open science commitment
R5Industrial partner disengagement or bankruptcyLowMediumWP1, WP5, WP7Secondment agreements with financial and IP terms settled at M3; alternative secondment hosts identified among associated partners; secondment obligations renegotiated if necessary, maintaining training objective
R6Data access or data licensing issues for AI trainingMediumHighScientific WPsData access agreements negotiated before DC recruitment; FAIR data commitments from data-providing partners formalised in Consortium Agreement; alternative publicly available datasets identified as fallback for each DC project
R7Compute resource bottlenecks for large-scale AI trainingMediumMediumScientific WPsAccess to EuroHPC compute time applied for in advance; institutional HPC resources at multiple beneficiaries provide redundancy; cloud compute budget reserved for burst capacity
R8IP disputes between academic and industrial partnersLowHighWP1, WP7IP provisions clearly defined in Consortium Agreement at M3; technology transfer offices of all beneficiaries engaged in drafting; pre-competitive research framing for most DC projects; any commercially sensitive DC projects assigned to industrial beneficiaries with clear IP terms
R9Open-source tools lack adoption by scientific communityMediumMediumWP7Early community engagement (workshops, tutorials) from M12; tools released in beta from M30 to gather user feedback before final release; integration with established domain-specific platforms and workflows; network with domain-specific user communities
R10Ethical concerns regarding AI use in sensitive scientific domains (e.g., medical AI, dual-use)Low–MediumHighWP1, WP5, WP6Ethics review at project start; ethics board includes external ethics expert; domain-specific risk assessments for DC-05, DC-11; training module on responsible AI (NWTE-6) addresses these proactively; alignment with EU AI Act requirements

3.2 Quality of consortium

Infrastructure and Capacity

[Write your content here]

[RAISE GUIDANCE: Describe the computational infrastructure available to the network for AI training and scientific computing: on-premise HPC clusters at beneficiary institutions, access to EuroHPC petascale systems, cloud agreements, and specialised AI hardware (GPU clusters, TPUs). Quantify available compute resources where possible (e.g., "INSTITUTION-X operates an 800-node GPU cluster providing approximately 10 petaFLOPS of AI training capacity"). Also describe domain-specific infrastructure: telescopes, synchrotrons, sequencing facilities, clinical data repositories, etc. that provide the scientific data for AI development. Infrastructure availability is a key differentiator for RAISE networks.**

Consortium Composition and Complementarities

[Write your content here]

[RAISE GUIDANCE: Explain why this specific combination of partners is optimal. Describe: (a) the scientific domain complementarities across the consortium — each partner brings unique domain expertise that, combined with the network's AI research, covers the scientific problem space comprehensively; (b) the AI methodology complementarities — different partners contribute different AI expertise (e.g., one partner leads in geometric deep learning, another in Bayesian methods, another in reinforcement learning); (c) the academic-industrial complementarity — academic partners drive foundational AI research while industrial partners provide validation environments, proprietary data, and translation expertise; (d) geographic complementarity ensuring the network serves a pan-European research community and contributes to reducing regional disparities in AI-in-science capacity. Include a brief narrative on how the consortium was assembled and why no critical competence is missing.**

Consortium Complementarity Matrix

PartnerPrimary AI ExpertisePrimary Domain ExpertiseUnique Infrastructure / ResourceDCs Supervised
[INSTITUTION-1][e.g., Deep learning, foundation models][e.g., Structural biology][e.g., 400-node GPU cluster, cryo-EM facility]DC-01, DC-05
[INSTITUTION-2][e.g., Scientific machine learning, physics-informed ML][e.g., Climate science, geophysics][e.g., Regional climate model infrastructure, EuroHPC allocation]DC-02, DC-13
[INSTITUTION-3][e.g., Generative models, Bayesian deep learning][e.g., Materials science, chemistry][e.g., High-throughput synthesis robot, DFT compute cluster]DC-03, DC-10
[INSTITUTION-4][e.g., Simulation-based inference, normalising flows][e.g., Astrophysics, HEP][e.g., Telescope data archive, particle physics simulation pipeline]DC-04, DC-07
[INSTITUTION-5][e.g., Multi-modal learning, self-supervised learning][e.g., Biomedical imaging, genomics][e.g., Clinical imaging dataset (10M images), sequencing core facility]DC-06, DC-08, DC-11
[INSTITUTION-6][e.g., Reinforcement learning, neural ODEs][e.g., Ecology, fluid dynamics][e.g., Long-term biodiversity monitoring datasets, CFD HPC facility]DC-09, DC-12
[COMPANY-1][e.g., Applied ML, drug design pipelines][e.g., Pharmaceutical R&D][e.g., Proprietary compound activity database, ADMET data, in-house wet lab]DC-05 (co), DC-14
[COMPANY-2][e.g., AI for Earth observation, data engineering][e.g., Satellite data, geospatial AI][e.g., Multi-year satellite imagery archive, GPU inference infrastructure]DC-09 (co), DC-15

Role of Associated Partners

[Write your content here]

[RAISE GUIDANCE: Each associated partner must have a clearly defined and substantive role — not merely a nominal one. For each, describe: (a) what specific expertise or resource they contribute; (b) which DCs will undertake secondments with them; (c) what training or scientific activities they participate in; (d) what bilateral agreements (Letter of Commitment, secondment agreements) underpin the relationship. Associated partners linked to a specific beneficiary should be explained in the context of that beneficiary's research activities.**

Funding of Non-Associated Third Countries (if applicable)

[Write your content here — or delete this subsection if not applicable]

[RAISE GUIDANCE: If any consortium member is from a non-EU, non-associated country, explain the scientific justification for their inclusion and confirm that their participation costs are borne by them or by a third-party funder, not by the MSCA grant. This section may be left blank or deleted if all participants are from EU Member States or Horizon Europe-associated countries.**


End of Part B1


Submission checklist (remove before submission):

- [ ] All [Write your content here] placeholders replaced with actual content

- [ ] All [PLACEHOLDER] values in tables replaced with real data

- [ ] All [RAISE GUIDANCE: ...] notes removed

- [ ] All institution names, supervisor names, and DC titles populated

- [ ] Table 3.1d: all 15 DC rows have correct duration values (3–36 months)

- [ ] Total body text (Sections 1–3) does not exceed 30 pages when formatted in the EC submission template

- [ ] All tables are properly formatted in the EC Word template (tables converted from Markdown)

- [ ] Part B2 (CVs, letters of commitment, ethics) completed separately

- [ ] Proposal submitted via HORIZON-MSCA-2026-DN-01-01 portal, tagged for RAISE call HORIZON-RAISE-2026-01-03

- [ ] Submission completed before 24 November 2026, 17:00 CET