Expert Consultancy on Machine Learning Integration for the Global Torture Index

The OMCT works with around 200 member organisations which constitute its SOS-Torture Network, to end torture, fight impunity and protect human rights defenders worldwide. Together, we make up the largest global group actively standing up to torture in more than 75 countries. Helping local voices be heard, we support our vital partners in the field and provide direct assistance to victims. Our international secretariat is based in Geneva, with offices in Brussels and Tunis.

ABOUT THE GLOBAL TORTURE INDEX

The objective of the Global Torture Index is to create a tool to measure and provide an assessment of the risk of being subjected to torture and other forms of ill-treatment in any given country on an annual basis. The measurement will assess State compliance with international standards on the fight against torture and ill-treatment both in laws and public policies but also in the practice.

The Global Torture Index will be launched in June 2025. Every year the Index will be launched with the analysis of the data conducted from the previous year. This initiative will primarily rely on the information provided by local actors, most of them being OMCT’s NGO partners and experts, as they have first-hand experience on the front line of the fight against torture in different contexts and in different parts of the world. This makes the Torture Index a member-driven initiative and thus the information will be collected and analysed at the local level with ownership, involvement and recognition of the implementing partners. The gathering of the information will be done through a detailed researched-driven and standarised questionnaire based on international and regional human rights standards.

ROLE & RESPONSIBILITIES

Overall Objective

The objective of this consultancy is to develop an efficient method to cross-check the Global Torture Index data using technology and/or machine learning. The process will aim to integrate various data sources to enhance the accuracy and reliability of the Index, ensuring consistency and credibility.

Specific Objectives

• To identify and evaluate the most effective technologies and methodologies for cross-checking Index data, focusing on technology, Artificial intelligence (AI), and machine learning solutions.
• To evaluate the feasibility of using machine learning and large language models to analyse data from the Index questionnaires across 26 countries, with a focus on identifying data inconsistencies.
• To develop a structured AI or machine learning framework for automating the cross-checking and validation process, ensuring methodological soundness and scalability, by:
  • Incorporating data from multiple sources, including those provided by anti-torture experts, civil society organisations, and human rights defenders working with the OMCT from the Year-1 Index countries, covering a total of 26 countries in the five regions, and amounting to approx. 75 organisations.
  • Cross-checking data with third-party reliable sources, including UN databases, independent civil society organisations, regional organisations, and national official data, to enhance the integrity of the Index data.
• To optimise AI-driven, machine learning, and other automated tools to improve the efficiency of cross-checking, filtering, and verifying the credibility of data.
• To provide insights and recommendations on the use of technology to enhance data credibility, automate workflows, and develop mechanisms to prove data reliability for trust-building in the tool.
• To ensure compliance with data protection, privacy regulations, anonymisation and ethical standards during the implementation of the cross-checking process, ensuring the do-no-harm principle.

TIMEFRAME, METHODOLOGY AND DELIVERABLES

Period covered: The consultancy will be conducted remotely throughout 2025, starting in May. The first phase from May to September will focus on more intensive work, including data analysis and technology assessments. The subsequent months will involve testing, refinement, and ongoing support extending through the remainder of the year.

Tasks and methodology: The consultant will undertake the following tasks:

1. Assess and Develop Methodology for Cross-Checking Index Data Using Technology

• Identify and evaluate potential technologies (AI, machine learning) to cross-check Index data.
• Assess the feasibility of using large language models to analyse data from Index questionnaires.

2. Develop AI and Machine Learning Frameworks for Data Validation

• Propose machine learning approaches for automating the cross-checking and verification of Index data.
  
• Integrating data from anti-torture experts, civil society organisations, and human rights defenders from the Year-1 Index countries (26).
• Compare the data from Index surveys with third-party sources, including UN databases, independent civil society organisations, regional organisations, and national official data.
• Develop algorithms or machine learning models to identify discrepancies and inconsistencies between the two above to improve the efficiency and accuracy of the data validation process.
  

3. Provide Recommendations for Enhancing Data Quality and Credibility

• Analyse how machine learning and AI can improve the reliability and credibility of data collected in the Index.
• Recommend tools and strategies for automating data filtering and improving information credibility.
• Address any potential issues related to data privacy and compliance with relevant regulations (e.g., GDPR).
• Testing phase and provide Ongoing Support to ensure effectiveness to the project.
• Offer continuous support and troubleshooting for the working model, including adjustments and fine-tuning.
• Provide recommendations for the future development of the Index data cross-checking process and for scaling the cross-checking system as well as integrating additional data sources.

Expected activities and key deliverables (non-exhaustive):

Task 1 Deliverables:

• Feasibility assessment report on technology options for data cross-checking, including AI and machine learning solutions (internal use).
• Documentation outlining the methodology for integrating technology into the data cross-checking process (internal use).
• Updated concept note detailing the methodology for using technology in Index data validation (public use).

Task 2 Deliverables:

• Integrated framework for cross-checking Index data with third-party sources, including sample comparisons (internal use).
• Technical documentation outlining the process of matching and validating data with third-party sources (internal use).
• Updated concept note section discussing third-party data integration (public use).

Task 3 Deliverables:

• Prototype machine learning models for automated data cross-checking and validation (internal use).
• Report on key findings, discrepancies, and patterns in the data, including recommendations for further validation (internal use).
• Participation in feedback sessions, meetings, and discussions about the validation process (internal and external use).

Task 4 Deliverables:

• Final report on the AI/machine learning framework, successful testing phase, including recommendations for its application in future Index rounds (internal and public use).
• Continued technical support for refining the cross-checking and validation process after the working model has been developed.
• Regular updates and consultations on improving data quality and ensuring consistency (internal use).

QUALIFICATIONS AND EXPERIENCE EXPECTED

Given the scale and complexity of the project, the consultancy may require collaboration with a research institution, university, or think tank to ensure the feasibility, sustainability, and academic rigor of the work. The consultant will serve as the primary focal point for the project, with the potential to collaborate with an affiliated institution or research entity to leverage additional expertise, resources, and institutional support—especially for the technical and methodological components of the data validation process.

The consultant should meet the following qualifications:

• University degree in data science, statistics, machine learning, or a related field. A postgraduate degree is an asset.
• At least 7 years of experience in data analysis and software engineering, with a strong focus on machine learning, back-end development, or cross-checking large datasets.
• Proven expertise in using statistical programming languages and machine learning frameworks for data processing and analysis.
• Experience with survey methodology and index construction, particularly in human rights, social sciences, or similar fields involving large-scale data is an asset.
• Strong understanding of data validation, cross-checking methodologies, and integrating multiple data sources (e.g., reports, databases, expert inputs) to ensure accuracy and consistency.
• Familiarity with human rights indicators, frameworks, and index-based analysis, including knowledge of data triangulation, scoring systems, and measuring human rights progress.
• Excellent oral and written communication skills in English, with the ability to clearly present technical concepts to non-technical stakeholders.
• Experience working with international organisations, CSOs, or research institutions on data-related projects, especially in the human rights or social justice sector is an asset.
• Knowledge of data protection and privacy regulations, as well as ethical considerations while handling sensitive data, including experience with GDPR or similar data protection frameworks.
• Ability to work independently in a remote setting, manage competing priorities, and meet deadlines while maintaining a high level of accuracy and attention to detail.

COMPETENCIES EXPECTED

• High ethical standards and proven commitment to human rights principles, data integrity, and research transparency.
• Strong organisational and project management skills, ensuring efficient and effective work delivery.
• Initiative and motivation: ability to work independently, take initiative, and drive tasks to completion.
• Advanced analytical and problem-solving capabilities to interpret complex datasets and derive meaningful insights.
• Familiarity with human rights measurement tools and methodologies, such as index scoring and comparative analysis.
• Commitment to human rights values and evidence-based policymaking.

EMPLOYMENT CONDITIONS

OMCT is an equal opportunities employer.
Duration of consultancy: May – December 2025.
Starting date: 1 May 2025
Place of work: Remote

ELIGIBILITY, APPLICATIONS, AND CONSIDERATION

Interested consultants are invited to submit their proposals via email to applications@omct.org with the subject line “Proposal for Global Torture Index Consultancy.” The submission should include the following attachments in PDF format:

• A Letter of Interest summarising your qualifications and interest in the project (maximum 1 page).
• A Bid for the Contract: detailing the following (maximum 2 pages):
  • Approach: Describe your proposed approach to the key tasks, including:
    • Refining and enhancing the data validation methodology, particularly with respect to the integration of AI and machine learning.
    • Cross-checking data between survey rounds and addressing key challenges such as inconsistencies and missing data.
  • Methodology: Highlight the tools, technologies, or machine learning frameworks you would employ, and explain how you will ensure the accuracy, reliability, and comparability of the data.
  • Timeline: Provide a rough breakdown of how you would approach the tasks and allocate time to each key deliverable. (e.g., Task 1: X weeks, Task 2: X weeks, Task 3: X weeks, etc.).
  • Collaboration: If applicable, briefly mention your planned collaboration with any institution, university, or think tank, and how you foresee leveraging additional expertise and resources to support the project.
  • Fee Structure: Indicate your proposed overall fee for the consultancy (taxes included), with any potential institutional or resource costs, if relevant. Provide a high-level breakdown of how you would allocate your time across the tasks.
• Curriculum Vitae (CV) or resume of the expert/s involved in the proposal.

Application Deadline: All proposals must be submitted by April 20th, 2025. Applications will be reviewed on a rolling basis, so early submission is encouraged.

The total budget allocated for the consultancy is approximately EUR 25’000. Competitive bids, including proposed methodologies, timeframes, and fee structures, will be given thorough consideration during the selection process.