Deliverables

Open source datasets suitable for semi-structured, unstructured and multi-turn conversations

Requirements for semi-structured, multi-turn and unstructured dialogues, available/relevant literature review on evaluation of dialogues and description of methodology on how the fused datasets are generated

Delivery Date: 31 Dec 2024

eMetrics for explainability and trustworthy responses

Methodologies and effectiveness in the context of semi-structured, unstructured, multi-turn dialogues. Codebase for metric implementation

Delivery Date: 31 Dec 2025

Methodology Assessment Framework

Finetuned methodology towards combined evaluation of LLMs in terms of its multimodality, multi-linguality, trustworthiness, explainability and understanding capabilities and its comparative analysis. Codebase for evaluation of dialogues.

Delivery Date: 30 Jun 2026

Conversational LLM

Methodologies for LLM finetuning and simulation/ augmentation of conversational training data, and applicability to use cases. Comparison with SOTA models. Codebase implementing the fine-tuning on selected datasets.

Delivery Date: 31 Dec 2024

Retrieval model for conversational style queries

The methodologies of integrating conversational LLM with FIR subsystem, discussion of open research questions in the context of ELOQUENCE. Codebase implementing the FIR, evaluation of the FIR precision given conversational queries.

Delivery Date: 30 Apr 2025

Methodology for jointly training FIR and response generator

The methodologies on interconnecting FIR and response generator, analysis of scenarios relevant to ELOQUENCE. Implementation of the FIR and response generator joint training.

Delivery Date: 30 Jun 2025

Interface for collecting user feedback

The definition of possible queries for user feedback, methodology for their generation, and their application during the dialog process. Implementation of the API for communicating feedback queries.

Delivery Date: 30 Sep 2025

Dialog system integrating components developed in T2.2-4

Description of the dialogue system architecture, interconnection of components, and training processes. Codebase for efficient training and testing of individual components in HPC environment, APIs for their interconnecting and joint training

Delivery Date: 31 Mar 2026

Algorithms definition, baselines, open issues and use cases

Methodologies applied for both knowledge-based approaches and semi-supervised learning, and applicability to use cases.

Delivery Date: 30 Jun 2025

Hybrid-AI and self-supervised models

The detailed definitions of the developed algorithms, code, and results from task T3.1 to T3.3. Releasing a codebase implementation.

Delivery Date: 28 Feb 2026

Interpretability and trustworthy AI

The intelligibility capabilities of models developed in T3.1-T3.3, and its impact on the user/agent satisfaction level.

Delivery Date: 31 Oct 2026

Definition of baselines, open issues, use cases scenarios and data

Methodology for identifying the baselines available in SOTA, the open issues to be specifically targeted given the use cases, the experimental/validation protocol and datasets.

Delivery Date: 31 Aug 2025

Multilingual, multimodal speech models

Deliverable on algorithms, recipes, code and an accompanying technical report related to the activities of tasks T4.1 and T4.2.

Delivery Date: 30 Apr 2026

Efficient, adaptable, bias-controlled, privacy-aware and ethical speech models

Deliverable on algorithms, recipes, code and an accompanying technical report related to the activities of tasks T4.3 and T4.4.

Delivery Date: 31 Dec 2026

Pilot Requirements & Usability Evaluation

Uses cases, requirements and outcomes for the different pilots. Summary of the set of KPIs, criteria and methodology to evaluate the ELOQUENCE usability.

Delivery Date: 31 Dec 2024

ELOQUENCE Interactive Playground

Initial version of the Interactive Playground for the LM validation.

Delivery Date: 30 Jun 2025

Pilots Integration

Integration according to the dialog flow of the different pilots

Delivery Date: 30 Apr 2026

Pilots Evaluation

Document summarising the pilot evaluations

Delivery Date: 31 Dec 2026

Report on linguistic expression respectful of EU values

Report on requirements for machine-generated verbal communication respectful of European values as enshrined in Article 2 of the EU Treaty.

Delivery Date: 30 Jun 2024

Emerging ELOQUENCE technology- approved by the ELOQUENCE Community

Finalized at TLR=3 stage, this report assesses emerging ELOQUENCE outputs as being respectful of EU values, with particular emphasis on gender, cultural or racial biases.

Delivery Date: 30 Jun 2025

ELOQUENCE societal acceptance and acceptability- approved by the ELOQUENCE Community

Finalized at TLR=4 stage, this report assesses ELOQUENCE outputs as being respectful of EU values, based on surveys amongst pertinent segments of society where there may be a heightened risk of communications being, or perceived as, discriminatory, biased or offensive.

Delivery Date: 31 Oct 2026

Final report on methodologies for assessing respect for EU values

Exportable lessons concerning the ethical, legal and societal assessment and validation of machine-generated verbal communication, with focus on methodologies and EU values.

Delivery Date: 30 Nov 2026

Dissemination, Communication and Exploitation Plan

Overall DEC plan, with KPIs and benchmarks, and campaign planning. Also includes the website.

Delivery Date: 31 Mar 2024

Dissemination, Communication and Exploitation Plan II

Overall DEC plan, with KPIs and benchmarks, and campaign planning. Also includes the website.

Delivery Date: 30 Jun 2025

Dissemination, Communication and Exploitation Plan III

Overall DEC plan, with KPIs and benchmarks, and campaign planning. Also includes the website.

Delivery Date: 31 Dec 2026

Business Case and Go-to-Market Assessment

Business case(s) and models for LLMs

Delivery Date: 31 Dec 2026

Standardisation Recommendations

Analysis and contributions to SDOs

Delivery Date: 31 Dec 2026

Project Management Handbook

Management procedure, consortium communication tools and the quality and risk management plans.

Delivery Date: 31 Mar 2024

Ethics Compliance Management Report

The Ethics requirements and compliance methodology for responsible research.

Delivery Date: 30 Apr 2025

Ethics Compliance Management Report II

The Ethics requirements and compliance methodology for responsible research.

Delivery Date: 31 Oct 2026

ELOQUENCE DMP

The legal aspects and the data management plan for the collection and processing of data throughout the project activities.

Delivery Date: 30 Jun 2024

ELOQUENCE DMP II

The legal aspects and the data management plan for the collection and processing of data throughout the project activities.

Delivery Date: 30 Jun 2025

ELOQUENCE DMP III

The legal aspects and the data management plan for the collection and processing of data throughout the project activities.

Delivery Date: 31 Dec 2026

OEI - Requirement No. 1

Ethics Advisor_A large quantity of online data, as well as voice data, will be processed in this project. Therefore, a data protection officer is required to assist in assessing the impact of the data processing on the rights of individuals and ensure that appropriate technical and organisational mitigation measures are in place from the start of the project (Privacy-by-design). The informed consent of human participants should also be documented. Safety and security risks, as well as the risks of misuse and manipulation of end-users via generative AI, should also be addressed by an Ethics Advisor. This could be the case, for example, in the event of secondary use of voice data in order to generate AI-based personalised content for advertisement purposes.

Delivery Date: 30 Apr 2024