Enhanced Technical Capabilities for EMSA Academy

Open tender by the European Maritime Safety Agency (EMSA), procedure EMSA/2026/OP/0011, to award a framework contract for AI-enabled tools and services for the EMSA Academy. Estimated maximum value €1,000,000 (excl. VAT) with contract duration up to 48 months and maximum annual maintenance cost €37,500. Scope includes design, development, model training, integration with MaKCs/VRESI and Moodle, demonstrations, training and operation under DevSecOps principles and compliance with the EU AI Act and data protection rules. Electronic submission via the EU Funding & Tenders Portal is required by 27 May 2026 16:00 Lisbon time; award will follow a best price-quality ratio and is open to economic operators established in EU Member States and specified neighbouring countries.

What it funds

Who can apply

A services framework tender to design, develop, train and operate AI-enabled tools for the EMSA Academy (AI4Academy). Work covers generative AI (including large language models), machine learning, agentic reasoning and related components to support course design, content generation, learning analytics, virtual scenarios, helpdesk automation, integration with Moodle (MaKCs) and VRESI, training and maintenance.

Contract form:Open procedure for a framework agreement (without reopening of competition). Award by best price-quality ratio. Electronic submission via the EU Funding & Tenders Portal e-Submission; EU Login is required ¹.

1. 1Service scope: analysis and design, AI model development and training, demos, integration into EMSA ICT, user training and support
2. 2Eligible applicants: economic operators established in EU Member States and listed neighbouring countries (see tender documents)
3. 3Core team profiles expected: Project Manager, AI/ML Architect, Data Scientist, ML Engineer, Data Engineer, Learning Technologist, Software Engineer, QA Officer

Contract value covers all services; pricing must be all-inclusive and fixed for the contract duration. Typical specific contracts (development, maintenance, operation) will be issued under the framework agreement. Tender documentation, templates and technical specifications are published on the F&T Portal.

Footnotes

1. 1Portal and e-Submission guidance: ec.europa.eu

Basic information and procurement context

Opportunity title:Enhanced Technical Capabilities for EMSA Academy (Procedure identifier: EMSA/2026/OP/0011). Lead contracting authority: European Maritime Safety Agency (EMSA). Procedure type: Open procedure for a services framework contract. Submission method: electronic submission through the EU Funding & Tenders Portal e-Submission (EU Login required). Estimated total value of the framework contract: €1,000,000 (excl. VAT). Maximum framework duration: 48 months. TED publication date: 21/04/2026. Deadlines: deadline for receipt of tenders 27/05/2026 16:00 Europe/Lisbon (submission by e-Submission); public opening 29/05/2026 10:00 Europe/Lisbon. Contract model: framework contract (FWC) without reopening of competition with specific contracts and order forms. Award method: best price-quality ratio. All submissions must follow Tender Specifications, Invitation to Tender and draft Framework Contract documents available on the F&T Portal.

Primary objective:Implement innovative AI-enabled tools to support and strengthen EMSA Academy’s design, development and delivery of learning services. The initiative is called AI4Academy and targets generative AI, large language models, machine learning, agentic reasoning and related techniques to improve personalised learning, training needs analysis, course design and content generation, simulation/VRESI integration, administration, reporting and user support.

Opportunity classification and key facts (answers to the required categories)

Eligible Applicant Types:Eligible applicants: private companies (startups, SMEs, large enterprises) providing IT and AI services, software development firms, system integrators; research institutions and universities (for consortiums); non-profit organisations/NGOs and training providers (where relevant); public-private consortia and joint offers are allowed. Subcontracting and reliance on other entities are allowed but must be declared and supported by commitment letters as per the Tender Specifications. International organisations may participate. Entities must not be subject to EU restrictive measures. Eligible tenderers include economic operators established in EU Member States and the following third countries: Albania, North Macedonia, Iceland, Liechtenstein, Montenegro, Norway, Kosovo, Serbia and Bosnia and Herzegovina.

Funding Type:This is a public procurement for services under a framework contract. The contracting mechanism is a procurement service contract (framework agreement) implemented through specific contracts and order forms. Not a grant, loan or equity instrument.

Consortium Requirement:Not mandatory to form a consortium. Single tenderers may apply. Joint offers (groups of economic operators) are permitted and, if used, group members are jointly and severally liable and must appoint a group leader. If a tenderer relies on other entities (subcontractors or entities on whose capacities they rely) these must be declared and supporting commitment letters provided for identified subcontractors and entities supplying selection criteria evidence.

Beneficiary Scope (Geographic Eligibility):Eligible geographic scope: entities established in EU Member States and economic operators established in eligible third countries listed in the Invitation to Tender: Albania, Republic of North Macedonia, Iceland, Liechtenstein, Montenegro, Norway, Kosovo, Serbia and Bosnia and Herzegovina. International organisations may also participate.

Target Sector:Target sectors and thematic scope: ICT, artificial intelligence, machine learning, generative AI, training technologies/learning technologies, simulation and virtual reality (VRESI), digital learning (Moodle/MaKCs), data engineering, cloud-native deployments, cyber security and compliance, public-sector digital services, adult and professional learning (maritime domain).

Mentioned Countries:Countries explicitly mentioned as eligible in the procurement: Albania; Republic of North Macedonia; Iceland; Liechtenstein; Montenegro; Norway; Kosovo; Serbia; Bosnia and Herzegovina; all EU Member States. EMSA also accepts international organisations. The contracting authority is EMSA (Lisbon, Portugal).

Project Stage (expected maturity):Expected maturity: development, integration, testing, demonstration and operation. The procurement covers analysis and design, model development and training, system demonstration (WP3), integration into EMSA ICT infrastructure (WP4), and maintenance, operation and training/support (WP5). Target TRL/TRL-like progress: applied development to demonstration and readiness for deployment and operations (validation and live operation).

Funding Amount:Estimated total value for the FWC: €1,000,000 (excluding VAT). Maximum price for one year of corrective maintenance and operation shall not exceed €37,500 (excluding VAT) as stated in the Tender Specifications.

Application Type:Open call for tenders (public) under the EU F&T Portal procurement module. Electronic submission only via e-Submission (EU Login account and, from June 30, 2026, two-factor authentication).

Nature of Support:Contracting authority will procure services. Beneficiaries (contractor) will receive monetary payment for services performed under specific contracts. The relationship is a commercial procurement contract; the outcome is services, deliverables, software, and operational support (monetary payments).

Application Stages:Single-stage tender submission for initial award of the FWC. Implementation proceeds through specific contracts. Evaluation includes exclusion and selection checks, technical capacity checklist, and award criteria evaluation (quality + price). Quality control and acceptance steps exist within project delivery lifecycle (multiple internal gates across PLAN → CODE → BUILD → TEST → RELEASE → DEPLOY → OPERATE).

Success Rates:Not provided by EMSA in the procurement documentation. Success rate depends on competition, compliance with selection criteria and passing minimum thresholds in award evaluation (quality minima per criterion and minimum overall score of 70%).

Co-funding Requirement:No co-funding from the tenderer is required by EMSA. This is a procurement: the contracting authority pays for services. Financial guarantees are not applicable except where stated in specific contracts; performance guarantees or other guarantees may be required as set out in the draft framework contract and specific contract templates.

Eligibility, selection and tender documentation

Key tender documents available on the F&T procurement page:Invitation to Tender; Tender Specifications (this document); Appendix TS.01 Price Summary Template; Appendix TS.02 Technical Capacity Checklist (with team profiles and project references worksheets); Appendix TS.03 Training Needs Analysis Procedure; Framework contract (IT) draft; General Terms and Conditions for IT contracts; Appendix A System and Application Technical Landscape; Appendix B Declaration and Conditions of Use for EMSA information systems; Appendix C DevSecOps Delivery (detailed process and deliverables); Appendix D Cloud Native Technical Landscape; Tender Q&A section (questions accepted until specified deadline). All supporting annexes and templates (price template, technical checklist, commitment letter templates, Legal Entities and Bank Accounts Form, Authorised Signatory Form, Declaration of Honour, etc.) must be completed and uploaded via e-Submission. Tender validity: 6 months.

1. 1Deadline for receipt of tenders: 27/05/2026 16:00 Europe/Lisbon (electronic via e-Submission).
2. 2Public opening of tenders: 29/05/2026 10:00 Europe/Lisbon (remote/public session).
3. 3Contract signature estimated: July 2026.
4. 4Framework contract maximum duration: 48 months.
5. 5Maximum framework budget: €1,000,000 (excluding VAT); maximum one-year corrective maintenance price: €37,500 (excluding VAT).

Language and submission:tenders may be submitted in any official EU language; English is preferred and working language of EMSA. Tenderers must register in the Participant Register and have a PIC. e-Submission imposes file size and format restrictions and accepts up to 200 attachments per submission. The PIC is mandatory. Tenderers should check the Tender Specifications section 13 (Submission via e-Submission) and EMSA procurement web pages for templates and instructions. All tenders must be signed by an authorised representative and include the completed Tenderer’s Checklist.

Selection and technical capacity requirements

Minimum economic/financial capacity:yearly turnover for last three years must be at least €300,000 (or equivalent evidence for public sector entities/individuals as specified). Grounds for exclusion: Declaration of Honour, judicial extracts or equivalent documents may be requested. Legal/regulatory and sanctions compliance checks apply (EU restrictive measures).

Technical and professional capacity (high-level summary):Tenderers must complete Appendix TS.02 Technical Capacity Checklist and provide between 3 and 5 reference projects from the last five years demonstrating: design and implementation of AI systems (ML/generative AI) across the full lifecycle; model training, testing and optimisation on real-world datasets; at least one project applying AI in learning/training/knowledge management contexts; integration into LMS or simulation environments; cloud-based production deployment and DevSecOps practices; delivery in public-sector or regulated environments; governance and compliance with AI/transparency/data protection requirements; delivery of training and user support. The tender must provide team CV summary and meet minimum profile requirements for Project Manager, AI/ML Architect, Data Scientist, Machine Learning Engineer, Data Engineer, Learning Technologist, Software Engineer and Quality Assurance Officer. At least one resource per profile must be proposed and satisfy education, experience and language requirements in Appendix TS.02.

Award criteria and scoring

The contract will be awarded to the most economically advantageous tender (highest score) based on weighted award criteria. Quality criteria weights and description are defined in section 16 of the Tender Specifications and include:

1. 1Quality criterion 1 (30%): AI-Driven Training Evaluation and Training Needs Analysis. Tenderers must propose solutions to convert qualitative feedback into KPIs, perform TNA and skills gap analysis, and describe data governance and GDPR compliance. Deliverables include conceptual analytics model, KPI examples, AI/NLP techniques and mock-up sample outputs.
2. 2Quality criterion 2 (20%): AI Helpdesk and Administrative Automation. Tenderers must detail chatbot/virtual assistant integration with MaKCs, escalation, logging, safeguards vs hallucinations, and integration with user management and reporting tools.
3. 3Quality criterion 3 (20%): Integration, Deployment and Operations. Tenderers must describe deployment in EMSA’s cloud-native environment, MaKCs/VRESI integration, DevSecOps practices, maintenance, incident management and documentation/knowledge transfer.
4. 4Price (30%): TOTAL price and Price breakdown to be submitted using Appendix TS.01 Price Summary Template. Price scoring uses lowest-price formula; total global score is quality+price.

Evaluation scoring:marks from 0–10 (half points allowed) for each quality criterion. Minimum thresholds apply: each quality sub-criterion must reach at least 60% (Q1, Q2, Q3 minima); only tenders achieving at least 70% overall score S will be eligible for award. Quality score SQ is weighted across criteria; price score SP calculated against lowest price. All scoring formulas are detailed in section 16 of the Tender Specifications.

Procurement process practicalities and obligations

Submission must be electronic via e-Submission (EU Login). Questions may be submitted through the F&T Portal Questions & Answers tab before the cut-off date. EMSA may not answer questions submitted less than six working days before the tender deadline. Tender validity period: 6 months. Tenderers may withdraw and replace tenders prior to the deadline; the latest tender counts. Tenderers are responsible for all costs and expenses for preparing and submitting tenders. PIC is mandatory. If submitting a joint tender the Agreement/Power of Attorney must be provided. Subcontracting rules: identified subcontractors above 20% of contract or on whose capacities tenderer relies must be listed and provide commitment letters (templates on EMSA procurement page). Entities on whose capacities tenderer relies (non-subcontractors) must provide commitment letters. EMSA reserves the right to reject or require replacement of subcontractors or entities in certain circumstances. Confidentiality and data protection: processing of personal data is governed by Regulation (EU) 2018/1725; tenderers must provide information required for evaluation and may be required to submit documentary evidence for Declaration of Honour reasons during evaluation. Participation is open on equal terms to economic operators from the countries listed and international organisations.

Technical and delivery templates and process guidance (how applications are assessed):Key templates and operational documents that bidders must use and follow: Appendix TS.01 Price Summary Template, Appendix TS.02 Technical Capacity Checklist (worksheets: Checklist, Tenderer’s Project References, Team CV Summary), Appendix TS.03 Training Needs Analysis Procedure, Invitation to Tender, Tender Specifications, Draft Framework Contract (IT), General Terms & Conditions for IT Contracts and Appendices A–D (System and Application Technical Landscape, Declaration and Conditions of Use for EMSA systems, DevSecOps Delivery, Cloud Native Landscape). DevSecOps Delivery Appendix C defines expected Plan→Code→Build→Test→Release→Deploy→Operate lifecycle tasks, deliverables, CI/CD and test automation integration, security assessment, and formal quality gates to be used for specific contracts. Source code delivery and versioning rules (GitLab, branch and release scheme), test automation and Jenkins/CI conventions and Flyway database migration usage are detailed in Appendix C and in the enclosed templates. Bidders must confirm they can operate within EMSA environment, use EMSA tools (Jira/Xray, Jenkins, Nexus, Confluence, Puppet/Flyway where applicable) or provide equivalent arrangements documented in the tender.

1. 1Mandatory deliverables in initial tender: completed Technical Capacity Checklist (Appendix TS.02), CVs (Europass format recommended) demonstrating required profiles, at least 3 project references (from at least 2 different clients) with description and links, signed cover letter and Authorised Signatory Form, Simplified Financial Statement, Declaration of Honour, Legal Entity and Bank Account form (unless already on file), Price Summary Template (Appendix TS.01), Tenderer’s Checklist signed.
2. 2Security and compliance obligations: designs must comply with the EU Artificial Intelligence Act (Regulation (EU) 2024/1689) as applicable; AI tools must be designed to avoid classification as high-risk systems for Education/Vocational training; data protection (GDPR / Regulation (EU) 2018/1725) compliance is mandatory when personal data processed; supplier to design for explainability and human oversight where appropriate.

Evaluation and award process:EMSA will carry out exclusion checks (Declaration of Honour), selection checks (legal, financial, technical capacity), compliance with minimum technical requirements, and award evaluation (quality + price). Documentation may be requested to verify Declaration of Honour items. The Technical Capacity Checklist in Appendix TS.02 is scored for essential selection conditions: number of AI projects, team profiles, and compliance with profile experience and education. The Agency reserves rights to cancel, negotiate special procedures for repetition of services up to 50% of initial contract value within three years (special negotiated procedure).

How to prepare a compliant proposal — practical checklist

Practical steps and minimum compliance items to submit a compliant tender:

1. 1Register organisation in Participant Register and obtain PIC; ensure EU Login account with 2FA available by June 30, 2026.
2. 2Download and use all procurement templates from EMSA procurement page (Invitation to Tender, Tender Specifications, Appendix TS.01, TS.02, TS.03, draft Framework Contract and its Appendices).
3. 3Complete Tenderer’s Checklist, Declaration of Honour, Legal Entities and Bank Accounts Form and Authorised Signatory Form.
4. 4Complete Appendix TS.02 Technical Capacity Checklist with project references and Team CV Summary. At least one resource per required profile must be included and meet the profile criteria.
5. 5Complete Appendix TS.01 Price Summary Template with day rates per profile and scenario calculations (Scenarios 1 and 2) and total price over 3 years including maintenance and annual presential meeting cost.
6. 6Prepare technical offer describing adherence to WP0–WP5 requirements, DevSecOps processes, AI governance (AI Act), ML lifecycle (data handling, training, validation, robustness, explainability), integration with MaKCs and VRESI, cloud-native deployment approach and operation in EMSA Landing Zone, security, and data protection safeguards.
7. 7Provide the required signed commitment letters from any identified subcontractors or entities on whose capacities the tenderer relies (if applicable) using EMSA model.
8. 8Submit the tender electronically via e-Submission before the deadline and ensure all mandatory fields and attachments are uploaded.

If you need EMSA procurement templates, guidance and the tender dossier visit the F&T Portal opportunity page and EMSA procurement pages. EMSA provides detailed tender specifications, an extensive set of appendices and templates and the DevSecOps Delivery appendix describing the quality gates and delivery obligations. All communications during the tender phase must be through the Questions & Answers tab of the F&T Portal. The contracting authority is not bound to reply to questions submitted after 19/05/2026 23:59 Europe/Lisbon as stated in the procurement notice.

Technical/operational expectations and mandatory rules for bidders

Mandatory technical expectations:proposals must include a DevSecOps-based delivery plan (Appendix C), evidence of ML development and operational experience (data pipelines, model lifecycle, performance monitoring, explainability, bias testing and governance), experience integrating AI with learning management systems (Moodle / MaKCs) and simulation environments (VRESI) using APIs and service-oriented architectures, cloud-native and Landing Zone deployment compatibility (Appendix D), compliance with AI Act and data protection rules, robust security and incident response procedures, and ability to deliver training / transfer-of-knowledge (WP5). Source code and documentation delivery, use of CI/CD and automated testing, and use of EMSA tools (Jira/Xray, GitLab, Jenkins, Nexus, Confluence, Puppet/Flyway) or documented equivalent are required. EMSA reserves the right to execute independent security assessments and penetration testing (CERT-EU and third-party assessments) prior to exposing systems to the internet and periodically thereafter; high-impact findings must be remediated immediately by the contractor.

AI governance and compliance:solutions involving AI must be designed so that EMSA can determine risk classification under the AI Act; EMSA expects designs that avoid the tool being classified as high-risk for Education and Vocational training systems, including controls for human oversight, explainability, accuracy, robustness, and data protection. Where personal data are processed, GDPR and Regulation (EU) 2018/1725 compliance is mandatory and appropriate data processing agreements, data minimisation, pseudonymisation and other safeguards shall be implemented.

Security and DevSecOps deliverables — summary:Appendix C DevSecOps Delivery is prescriptive: PLAN→CODE→BUILD→TEST→RELEASE→DEPLOY→OPERATE loops, automated CI/CD, static and dynamic security testing (SAST/DAST), test automation (ReadyAPI, JMeter, Cucumber with Cypress/Selenium), source code analysis (Sonar, Checkmarx/Veracode equivalent), infrastructure deployment automation (Puppet modules, Flyway database migrations), artifact management (Nexus), test and deployment pipelines (Jenkins) and log/monitoring integration (Splunk, Prometheus/Grafana, Azure Monitor, Nagios). The contractor must produce and maintain the listed deliverables: architecture documentation, technical sheets, dependency lists, datasets descriptors, training materials, test plans and results, CI/CD scripts, source code in EMSA GitLab with the prescribed branching and release scheme, vulnerability remediation logs and incident response procedures. EMSA will perform formal quality control gates and may withhold acceptance if gates are not satisfied.

How we explain this opportunity in a single paragraph

This is a competitive public procurement by EMSA to award a framework contract (services) for the AI4Academy programme: a multi-year technical partnership to design, develop, integrate, deploy and operate AI-enabled tools that enhance EMSA Academy learning services. The work spans analysis and design, model development and training, demonstration, secure integration with EMSA platforms (MaKCs, VRESI), cloud-native deployment and long-term operation and maintenance under a DevSecOps approach. Tenderers must demonstrate strong multidisciplinary experience across AI/ML engineering, learning technologies, cloud/DevSecOps operations, data engineering, security and public-sector delivery. Proposals must comply with the Tender Specifications, appendices and templates, use the e-Submission system, satisfy minimum selection criteria including turnover and project references, meet profile and technical checklist requirements and comply with AI Act and data protection obligations. The maximum available estimate for the framework is €1,000,000 and the procurement is open to eligible economic operators located in the EU and the listed neighbouring countries.

Full tender dossier and updates are available on the F&T Portal and on EMSA procurement pages. For direct access to the opportunity see the tender details on the F&T Portal EMSA Tender EMSA/2026/OP/0011 ¹

Footnotes

1. 1Procurement documentation, appendices, templates and submission via the Funding & Tenders Portal (EU F&T Portal): ec.europa.eu

Opportunity Overview

This is an open tender procedure launched by the European Maritime Safety Agency (EMSA) to procure innovative AI-enabled tools and services for the EMSA Academy. The contract aims to implement generative AI, Large Language Models, Machine Learning, and Agentic Reasoning technologies to enhance the design, development, and delivery of learning services for maritime professionals and beneficiary organizations across EU Member States and partner countries.

Key Funding Details

Procedure Identifier:EMSA/2026/OP/0011

Procedure Type:Open procedure with publication in the Official Journal (TED reference 77/2026 271816-2026)

Contract Type:Framework Contract for IT Services (single FWC without reopening of competition)

Estimated Total Budget:€1,000,000.00 (excluding VAT). This represents the maximum amount available; EMSA is not obligated to spend the full amount.

Maximum Annual Maintenance Cost:€37,500.00 (excluding VAT) per year for corrective maintenance and operation services

Contract Duration:48 months from the date of entry into force, with an estimated signature date of July 2026. The FWC may be renewed once for maintenance only for a maximum of 12 months with express written agreement of both parties.

Critical Deadlines

Tenders submitted after the deadline or through any means other than electronic submission via the e-Submission application on the Funding and Tenders Portal will be rejected. The tender validity period is 6 months from submission.

Scope of Work and Deliverables

The contract encompasses the development and implementation of AI-enabled tools across six Work Packages (WPs), structured as follows:

• WP0: Project Management and Reporting - including AI system architecture documentation, technical sheets, library and dependency documentation, and final project reports
• WP1: Analysis and Design of AI-Integrated Tools - tool conceptualization, stakeholder feedback gathering, feasibility studies, and implementation roadmaps
• WP2: Development and Training of AI Systems - data collection and preparation, AI model selection and development, training and testing in simulated environments, and model refinement with continuous learning mechanisms
• WP3: Demo of the Developed System - final demonstration to stakeholders and live validation sessions confirming system compliance with specifications
• WP4: Integration into EMSA ICT Infrastructure - deployment on EMSA cloud infrastructure (standalone, MaKCs, or VRESI integration), development and enhancement services, and corrective maintenance and operation under defined SLAs
• WP5: Training and Support - comprehensive training courses for EMSA staff in English (in-person, online, or eLearning), alternative training modalities (webinars, ad-hoc sessions), and user support services

Specific Use Cases and Requirements

The tender specifications detail three primary quality criteria that will be evaluated:

Quality Criterion 1: AI-Driven Training Evaluation and Needs Analysis (30% weighting):Tenderers must describe how their solution collects and processes qualitative and quantitative learning data, converts free-text feedback into measurable KPIs, identifies trends and improvement actions, and supports prospective Training Needs Analysis (TNA) and skills gap analysis. Applicants must provide a conceptual analytics model, examples of KPIs and dashboards, descriptions of AI/NLP techniques, sample output reports, data governance measures, and training support approaches.

Quality Criterion 2: AI Helpdesk and Administrative Automation (20% weighting):Tenderers must describe how their AI solution acts as a chatbot or virtual assistant for common administrative queries, integrates with MaKCs user management and reporting tools, escalates complex cases to human operators, and produces real-time statistics and logs. Required evidence includes use-case flow diagrams, example dialogue scenarios, integration point descriptions, safeguards against hallucinations, monitoring and logging approaches, and staff training elements.

Quality Criterion 3: Integration, Deployment, and Operations (20% weighting):Tenderers must describe deployment in EMSA's cloud-native ICT environment, integration with MaKCs and/or VRESI, DevSecOps practices and maintenance approaches aligned with SLA requirements, and documentation and knowledge transfer strategies. Required evidence includes deployment and integration plans aligned with EMSA Landing Zone principles, DevSecOps and quality assurance process descriptions, maintenance and operation approaches, and sample technical and operational documentation structures.

Price Criterion (30% weighting):Prices are evaluated based on best price-quality ratio methodology. The tender must include pricing for two development scenarios: Scenario 1 (Development of a learning AI-enabled assistant using retrieval-augmented generation) and Scenario 2 (Development of an AI-enabled learning contents editor for Moodle and H5P packages), plus 3 years of maintenance costs for each scenario.

Eligibility and Selection Criteria

Geographic Eligibility

In addition to EU Member States, tenderers from the following countries are eligible:Albania, Republic of North Macedonia, Iceland, Liechtenstein, Montenegro, Norway, Kosovo, Serbia, and Bosnia and Herzegovina.

Economic and Financial Capacity

Tenderers must demonstrate a minimum yearly turnover of €300,000.00 for the last 3 years. Evidence must include profit and loss accounts, balance sheets, or equivalent financial statements depending on the legal form of the organization (profit, non-profit, public sector, or individual).

Technical and Professional Capacity

Tenderers must demonstrate strong, multidisciplinary experience at the intersection of advanced AI technologies, digital learning systems, and organizational transformation. This includes proven expertise in generative AI, Large Language Models, machine learning, and agent-based systems; demonstrated experience in learning technologies and educational innovation in adult learning and professional training contexts; strong background in learning platforms and simulations with AI integration experience; experience in skills analytics and workforce intelligence; capability to design AI-supported collaboration and community learning tools; experience in complex, regulated, or public-sector environments; and proven ability to integrate new AI solutions with existing systems. ¹

Tenderers must provide evidence through a list of 3 to 5 projects or contracts completed in the last 5 years from at least 2 different clients. Projects must have been executed as sole contractor, group member (with at least 15% of total value), or subcontractor. Framework contracts with multiple specific contracts are considered a single reference. Different departments of the same body are regarded as the same client.

Required Team Structure

Tenderers must propose a minimum team consisting of the following profiles, each with specific education, English language proficiency (minimum B2 level), and years of relevant experience requirements:

• Project Manager (PM) - minimum 5 years managing IT or software development projects, with experience in complex multi-stakeholder projects and multidisciplinary team coordination
• AI/ML Architect (AMA) - minimum 5 years in software architecture design with at least 3 years involving AI/ML systems, experience designing end-to-end AI system architectures and working with machine learning and generative AI frameworks
• Data Scientist (DS) - minimum 3 years as Data Scientist or similar role, with experience in data analysis, feature engineering, model evaluation, and working with real-world datasets
• Machine Learning Engineer (MLE) - minimum 5 years implementing and deploying machine learning models, with experience in model training, testing, optimization, and integrating ML components into software systems
• Data Engineer (DE) - minimum 5 years in data engineering or data platform development, with experience designing and implementing data pipelines and supporting AI/ML workloads
• Learning Technologist (LT) - minimum 5 years in learning technology, instructional design, or digital learning development, with experience designing training content for adult and professional learning contexts and working with Learning Management Systems (e.g., Moodle 4.5 or equivalent)
• Software Engineer (SE) - minimum 5 years in software development with at least 3 years developing backend and/or frontend components for enterprise or web-based applications, experience integrating external services via APIs and working in DevSecOps and CI/CD environments
• Quality Assurance Officer (QAO) - minimum 3 years in software quality assurance or testing, with experience defining and executing test plans and validating functional and non-functional requirements

Tenderers may nominate multiple resources for each profile to ensure comprehensive technology coverage. Every proposed resource must independently satisfy all stated criteria, including university degree requirements (or equivalent professional experience of at least 6 years), English proficiency at B2 level or higher, and specified years of relevant experience. ²

Exclusion and Conflict of Interest Requirements

Tenderers, group members, subcontractors, and entities on whose capacities the tenderer relies must not be subject to EU restrictive measures or in any exclusion situation as defined in Article 138 of Regulation (EU, Euratom) No 2024/2509. Tenderers must complete and sign a Declaration of Honour confirming non-exclusion. Tenderers must also ensure they are not in situations of conflict of interest or professional conflicting interest that could compromise impartial and objective implementation of the contract.

Submission Requirements and Process

All tenders must be submitted electronically via the e-Submission application accessible from the Funding and Tenders Portal. Tenders submitted by any other means (email, hard copy, etc.) will be rejected. Tenderers must have an EU Login account to access e-Submission and submit questions through the Questions and Answers tab.

Required submission documents include:a signed cover letter with authorized signatory information and bank account details; the Legal Entities and Bank Accounts Form (if not previously submitted to EMSA); evidence of legal and regulatory capacity; evidence of economic and financial capacity; evidence of technical and professional capacity including project references and team CVs; all information required for award criteria evaluation; and pricing information following the Price Summary Template (Appendix TS.01).

The Technical Capacity Checklist (Appendix TS.02) must be completed with detailed information on organizational criteria relating to the tenderer and individual criteria for each proposed team member. Team members must provide Europass CVs highlighting information corresponding to each profile criterion.

Evaluation and Award Methodology

The contract will be awarded to the tenderer submitting the most economically advantageous tender based on a best price-quality ratio approach. Evaluation follows a two-stage process:

1. 1Compliance verification: checking EU restrictive measures, access to procurement, non-exclusion status, absence of professional conflicting interests, selection criteria compliance, and minimum requirements verification
2. 2Award criteria evaluation: scoring tenders on three quality criteria (each worth 30%, 20%, and 20% respectively) and price (30%). Evaluators assign marks between 0-10 for each quality criterion. Only tenders achieving minimum 60% on each of the three quality criteria and minimum 70% overall score are considered for award.

The scoring formula is:S = SQ + SP, where SQ is the weighted quality score and SP is the price score. Quality scores are calculated as the average of evaluator marks multiplied by criterion weightings. Price scores are calculated based on the ratio of the lowest price to each tender's price, multiplied by the price weighting.

Contract Terms and Conditions

Payment Terms:Payments are made in euros following acceptance of deliverables. For development and enhancement services, interim payments may be made at agreed milestones (typically 30% at 4 months, 30% at 8 months, and final payment at contract end). For maintenance and operation services, typical payment structure includes interim payments at 6-month intervals (30% each) and final payment at contract end. All invoices must be submitted within 60 days of service completion and include clear descriptions of services provided and contract references. The contracting authority has 60 days to process and pay invoices from receipt.

Intellectual Property Rights:The Union acquires irrevocably worldwide ownership of all results and intellectual property rights created under the contract. Pre-existing rights are licensed to the contracting authority on a royalty-free, non-exclusive, irrevocable basis for all modes of exploitation. The contractor must provide a complete list of pre-existing rights and evidence of ownership or licensing authority. All results are considered work for hire under Portuguese law, with all associated intellectual property rights vesting in the contracting authority.

Data Protection and Security:Processing of personal data must comply with Regulation (EU) 2018/1725 and Regulation (EU) 2016/679. Personal data shall only be processed within the EU and EEA territories and held in data centers located within these territories. No access to personal data is permitted outside the EU and EEA. The contractor must implement appropriate technical and organizational security measures, maintain records of data processing operations, and notify the contracting authority of any personal data breaches within 48 hours. ³

Artificial Intelligence Act Compliance:All AI capabilities and techniques must comply with Regulation (EU) 2024/1689 (Artificial Intelligence Act). Tools developed under the contract must not be qualified as high-risk in accordance with the AI Act. Special attention must be devoted to assessing interpretability of AI techniques. The contractor must design solutions ensuring explainability of AI outputs, human oversight mechanisms, and data protection safeguards. EMSA retains final determination of AI risk classification. Notably, EMSA Academy's AI tools shall not affect, influence, or dictate access or steering of individual learners' educational and professional courses, as this remains the prerogative of relevant competent authorities of Member States and beneficiary countries.

Subcontracting:Subcontracting is permitted with prior written authorization from EMSA. Tenderers must identify subcontractors whose individual share exceeds 20% or on whose capacities the tenderer relies to fulfill selection criteria. Identified subcontractors must provide commitment letters. The contractor remains fully liable for contract performance regardless of subcontracting arrangements. Second-level subcontracting is not admissible except with prior written authorization.

Termination and Liability:Either party may terminate the contract for convenience with 9 months written notice. EMSA may terminate for cause including failure to implement the contract in accordance with specifications, material breach of obligations, repeated refusal to sign specific contracts, contractor's exclusion situations, conflicts of interest, or force majeure. The contractor's liability is limited to three times the value of the relevant specific contract, except in cases of gross negligence, willful misconduct, injury to persons, or third-party intellectual property claims. Liquidated damages for delay are calculated as 0.3 x (V/d), where V is the contract value and d is the duration in days.

Specific Conditions and Constraints

Prices must be fixed, non-revisable amounts in euros, all-inclusive of all costs associated with service provision including travel. EMSA is exempt from all taxes and duties including VAT under Protocol 7 of the Treaty on the Functioning of the European Union. The contractor must complete necessary formalities to ensure tax exemption. Prices must address all elements specified in the Price Summary Template (Appendix TS.01), including person-day rates for each profile and annual maintenance costs.

The contractor must ensure personnel implementing the contract possess required professional qualifications and experience as specified in the selection criteria. At EMSA's reasoned request, the contractor must replace any personnel lacking required expertise or causing disruption at EMSA premises, bearing all replacement costs. The contractor must comply with EMSA's internal security rules and code of conduct requirements for contractor staff.

All services must be delivered in English, the working language of EMSA. In-person training courses must be held at EMSA premises in Lisbon. Online courses must be delivered via Microsoft Teams or Zoom. eLearning material must be delivered via MaKCs using native Moodle activities and resources, H5P interactive contents, or multimedia products. One annual presential meeting covering all active services must be included in the price offer.

The contractor must maintain all original documents for 5 years from payment of the final invoice. EMSA may conduct checks and audits at any time during service provision and up to 5 years after final payment. The contractor must grant EMSA staff and authorized external personnel appropriate access to sites, premises, and all information needed for audits and checks.

Key Considerations for Applicants

Applicants should note that this is a framework contract with implementation through specific contracts. EMSA is not obligated to order services for the maximum amount. The contract structure allows for flexibility in scope and timing of specific contracts, which will be determined during implementation based on EMSA's evolving needs. Applicants must demonstrate not only technical capability in AI and machine learning but also deep understanding of adult learning, maritime training contexts, and organizational change management. The evaluation heavily emphasizes practical experience with similar projects and the ability to deliver integrated solutions that enhance rather than replace existing EMSA systems and practices. Applicants should carefully review all tender specifications, appendices, and draft contract terms before submission, as these documents contain detailed technical requirements, quality standards, and contractual obligations that will govern the relationship. Questions regarding the tender should be submitted through the F&T Portal by the specified deadline; EMSA is not bound to reply to questions submitted less than 6 working days before the submission deadline.

Footnotes

1. 1Detailed selection criteria are specified in Section 15.5 of the Tender Specifications. Tenderers must provide evidence through project references demonstrating experience in AI system development, learning technology implementation, and integration with existing platforms.
2. 2The Technical Capacity Checklist (Appendix TS.02) provides detailed requirements for each profile including specific experience criteria, education requirements, and English language proficiency levels. Tenderers must complete this checklist comprehensively and provide supporting Europass CVs for all proposed team members.
3. 3Article I.9.2 of the Framework Contract specifies that personal data shall only be processed within the territory of the European Union and the European Economic Area, with data held exclusively in data centers located within these territories. The contractor may not change the location of data processing without prior written authorization from EMSA.