The ultimate result of the service will be the delivery of a fully functional INFF-AI pilot, developed, tested, and deployed within a UN-approved environment to support countries in conducting financing assessments and preparing financing strategies in line with INFF guidance. The tool will enable users to upload national plans, budgets, and relevant financing documents to generate structured financing assessments, draft outlines of financing strategies guided by a decision-tree, and “INFF Check” analyses of sectoral or thematic strategies, highlighting gaps and suggesting INFF-aligned improvements. Customized for at least two pilot countries, the solution will be developed primarily on Microsoft Azure services—or, if Azure proves unsuitable, on an alternative platform meeting all UN OICT requirements. Following successful OICT security clearance, the tool will be deployed to production, accompanied by complete technical documentation, source code, user guides, and training materials. User testing with government counterparts and UN country teams will inform final refinements, and recommendations will be provided for scale-up, sustainability, and integration into UNDESA’s broader support tools.

Home-Based

3 Months

The Integrated National Financing Framework (INFF) provides a country-led approach to strengthen planning, financing, and implementation of national sustainable development strategies. As countries seek to translate national development plans into actionable and resourced strategies, there is growing demand for technical support, including assistance in conducting financing assessments and outlining potential elements for financing strategies. To help address this need and improve the speed, consistency, and accessibility of INFF-related support, the United Nations Department of Economic and Social Affairs (UNDESA) is exploring the development of “INFF-AI”—a generative artificial intelligence assistant designed to support key stages of the INFF process. This pilot initiative will develop an AI-based tool to assist countries in: • Conducting a Financing Assessment of their national development plans • Laying out potential components of a Financing Strategy, aligned with INFF guidance • Conducting an “INFF Check” for existing sectoral or thematic financing strategies The objective of this consultancy is to design, develop, and deploy a pilot version of INFF-AI based on uploaded national plans/budgets, existing financing strategies, relevant financing data or other user inputs that can: • Guide users through a structured financing assessment of a country’s national plan based on INFF building block 1 methodology • Generate a draft outline of potential elements for a financing strategy based on INFF building block 2 methodology, incorporating a decision-tree to guide users through strategic choices • Analyse existing financing strategy documents (e.g., climate finance strategy, health finance strategy, disaster risk reduction finance strategy) and generate suggested INFF-aligned improvements or next steps • Be tailored for selected pilot countries (e.g., Dominican Republic, Seychelles, Fiji) Work Assignments: Under the supervision of the Inter-Regional Adviser of UNDESA’s Financing for Sustainable Development Office, the consultant(s) will: Phase 1: Design • Review existing INFF guidance, templates, existing financing strategies, and country case studies • Define user personas (e.g., government planner, UN adviser, consultant) • Design AI-assisted workflows for: o Financing assessment of national development priorities o Drafting outline of potential elements for national financing strategies o Developing and integrating “INFF check” function • Identify and assess suitable AI tools and hosting platforms, prioritizing the trial use of Microsoft Azure services approved by use by the UN Office of Information and Communications Technology (OICT). If Azure proves unsuitable for the intended functionalities or pilot use cases, alternative platforms may be proposed – provided they meet UN OICT requirements (see Section 4) Phase 2: Development • Develop the pilot version of INFF-AI, including: o Natural language interface with tailored prompt structures o Upload capability for national development plans, financing strategies, and financing data (as feasible) o Output generation for analytical summaries, financing matrices, and draft text o Incorporate a decision-tree element to guide users through structured choices based on INFF Building Block 2 methodology, enabling users to:  Identify strategic financing objectives  Choose relevant financing policy areas and instruments  Sequence reforms and prioritize actions o Develop and integrate an “INFF Check” function to allow users to upload existing financing or policy documents (e.g., climate finance strategy, health finance strategy, disaster risk reduction finance strategy) and receive:  Highlights of potential gaps, overlaps, or areas for further integration/development,  Suggested INFF-aligned improvements or next steps. • Customize the tool for at least two pilot countries Phase 3: Deployment • Deploy the final product on the development environment for OICT to conduct the security test • Once the final product passes the OICT’s security test, deploy it to the production environment. Phase 4: Testing and Refinement • Conduct user testing with government officials and UN country teams • Incorporate feedback to improve usability and accuracy • Produce a user guide and training material Phase 5: Documentation and Handover • Provide source code/scripts (if applicable) • Document technical architecture and training data sources • Recommend options for scale-up, sustainability, and integration into UNDESA support tools Technology Options and Hosting Considerations: The consultant will assess the most appropriate platform for deploying INFF-AI based on cost, usability, data governance, and scalability. As a first step, the tool should be developed and trialed using Microsoft Azure services, which are approved for use by UN OICT and already available to DESA. The preferred options include: • Azure OpenAI – Access to GPT models (e.g., GPT-4) via Azure with enterprise-grade security; enables secure, UN-compliant deployments with usage-based pricing. • Azure Foundry models – Proprietary Microsoft-hosted models for various AI tasks; integrated with other Azure AI services. • Copilot Studio – Low-code platform to build, customize, and publish AI copilots integrated with Microsoft 365 apps and Teams workflows; ideal for rapid development by business users but offers limited flexibility for advanced LLM use cases. • AI Hub in Power Apps – Central hub in Power Platform to use prebuilt or custom AI models within apps, including integration with Azure OpenAI; best for automation and structured tasks, with limited LLM support and not suited for open-ended generation. • Other Azure portal services – Full suite of AI tools such as Azure ML, Cognitive Services, and Data Lake for custom AI model deployment and orchestration; offers high flexibility and scalability but requires strong technical expertise and infrastructure planning. If Azure proves unsuitable for the required functionality or integration, the consultant may recommend alternative platforms. These may include commercial APIs (e.g., OpenAI GPT-4 Turbo, Anthropic Claude), open-source models (e.g., LLaMA 3, Mistral), or tools developed with UN partners (e.g., UNICC, UN Global Pulse). Any recommendation should be based on a clear assessment of feasibility, data security, sustainability, and alignment with digital public goods principles.

• An advanced degree (Master’s or higher) in Computer Science, Artificial Intelligence, Data Science, Software Engineering, or a related field. A relevant first-level university degree with two additional years of qualifying experience may be accepted in lieu of the advanced degree. • Minimum of 5 years’ experience in AI/ML application development, including demonstrated expertise with large language models (LLMs) such as GPT (via Azure OpenAI), Claude, or open-source models (e.g., LLaMA). • Experience developing AI-driven decision-support tools or knowledge-based systems (e.g., decision trees, workflow engines, rule-based interfaces). • Proven experience working with Microsoft Azure AI services (e.g., Azure OpenAI, Copilot Studio) or ability to quickly adapt to the Azure ecosystem. • Familiarity with issues related to data governance, hosting, and deployment in cloud-based environments. • Understanding of public financial management, development planning, or sustainable development finance. • Prior experience working with international organizations or in development contexts is desirable. • Excellent communication and documentation skills, and ability to collaborate effectively in a remote setting.

English and French are the working languages of the United Nations Secretariat. For this post, fluency in English is required. Knowledge of another UN official language is desirable.

Not available.

THE UNITED NATIONS DOES NOT CHARGE A FEE AT ANY STAGE OF THE RECRUITMENT PROCESS (APPLICATION, INTERVIEW MEETING, PROCESSING, OR TRAINING). THE UNITED NATIONS DOES NOT CONCERN ITSELF WITH INFORMATION ON APPLICANTS’ BANK ACCOUNTS.