IT Officer, Business Solutions I (Agentic AI Developer)

Do you want to build a career that is truly worthwhile? Working at the World Bank Group provides a unique opportunity for you to help our clients solve their greatest development challenges. The World Bank Group is one of the largest sources of funding and knowledge for developing countries; a unique global partnership of five institutions dedicated to ending extreme poverty, increasing shared prosperity and promoting sustainable development. With 189 member countries and more than 130 offices worldwide, we work with public and private sector partners, investing in groundbreaking projects and using data, research, and technology to develop solutions to the most urgent global challenges. For more information, visit www.worldbank.org ITS Vice Presidency Context The Information and Technology Solutions (ITS) Vice Presidential Unit (VPU) enables the World Bank Group to achieve its mission of ending extreme poverty and boost shared prosperity on a livable planet by delivering transformative information and technologies to its staff working in over 150+ locations. For more information on ITS, see this video: https://www.youtube.com/watch?reload=9 v=VTFGffa1Y7w Background The World Bank Group (WBG) is accelerating its use of artificial intelligence to enhance operational efficiency, knowledge management, and development impact. As part of this effort, the Information Technology Solutions (ITS) vice presidency — and specifically the ITSSG team — is expanding its capacity to design, build, and deploy agentic AI systems: autonomous, multi-step AI workflows capable of reasoning, tool use, planning, and decision-making with minimal human intervention. This Terms of Reference outlines the scope of work, responsibilities, and qualifications for an Agentic AI Developer at the GF staff level, based in Sofia, Bulgaria, to support ITSSG's AI platform and product development initiatives. Objective The Agentic AI Developer will design and implement production-grade agentic AI systems that integrate large language models (LLMs) with tools, data sources, and enterprise workflows. The staff member will work closely with the ITSSG team, product managers, and business units to translate operational needs into scalable AI-powered solutions. Scope of Work The staff will be responsible for the following: 1. Agentic System Design and Development • Architect and implement multi-agent and single-agent AI systems using frameworks such as LangGraph, LangChain, AutoGen, CrewAI, or equivalent. • Design agent reasoning loops, tool-calling mechanisms, memory management strategies, and orchestration logic. • Develop agents capable of interacting with APIs, databases, document repositories, and internal WBG systems. • Implement retrieval-augmented generation (RAG) pipelines to ground agents in institutional knowledge and internal data. 2. LLM Integration and Prompt Engineering • Integrate leading LLM providers (Anthropic Claude, OpenAI GPT, Google Gemini) into agent workflows via APIs. • Design and refine system prompts, few-shot examples, and chain-of-thought instructions to guide agent behavior reliably. • Implement guardrails, output validation, and fallback logic to ensure safe and accurate agent responses. 3. Tool and API Development • Build custom tools and function-calling interfaces that agents can invoke (e.g., search tools, database queries, document processors, email senders). • Ensure tool schemas are well-defined, type-safe, and robust against edge cases. • Integrate with internal WBG platforms and data sources as required. 4. Evaluation and Testing • Design evaluation frameworks to assess agent performance, accuracy, groundedness, and safety. • Conduct red-teaming and adversarial testing to identify failure modes and jailbreak risks. • Establish benchmark datasets and automated testing pipelines for regression monitoring. 5. Deployment and MLOps • Package and deploy agentic applications on cloud infrastructure (Azure, GCP, or equivalent). • Implement observability and monitoring for agent behavior, latency, and token usage. • Maintain CI/CD pipelines for continuous delivery of agent updates. 6. Documentation and Knowledge Transfer • Produce technical documentation covering architecture decisions, agent workflows, and API specifications. • Provide knowledge transfer sessions and code walkthroughs for internal engineering teams. • Contribute to internal guidelines and best practices for agentic AI development at WBG. Deliverables The staff member will be expected to contribute to ongoing Agentic AI initiatives and will set annual deliverables in alignment with the WBG performance and talent management framework. Key expected outputs include: • Architecture and design documentation for assigned agentic systems. • Functional, tested, and deployed agentic AI solutions integrated with WBG platforms. • Evaluation and monitoring reports covering agent performance, safety, and reliability. • Technical documentation and active knowledge sharing within the team.