AI Engineer: LLM & Agent Solutions for Global Impact

World Bank Group; Washington; On-site; USD 120,000 - 160,000; Full time

Job summary

The World Bank Group is seeking an ET Consultant 1 – AI Engineer to design, develop, and deploy AI agent systems that integrate with Azure, Databricks, and internal APIs. You will build LLM-enabled features, orchestrate multi-step workflows, and implement responsible AI guardrails.

Strong collaboration across cross-functional teams is essential. You will also produce architecture docs, runbooks, and training materials while ensuring security, data governance, and scalable design in a regulated,

Qualifications

• Bachelor's degree in a relevant field (preferred: Master's).
• 2+ years of AI Engineer experience with AI/ML/LLMs.
• Hands-on experience designing and developing AI solutions (LLMs, RAG, AI systems).
• Experience with LLM APIs, vector search, tool calling, context engineering, and prompt design.

Responsibilities

• Develop AI solutions with LLMs, prompts, tools, and orchestration.
• Build and tune RAG pipelines (ingestion, embedding, retrieval) and evaluations.
• Deploy and operate solutions on Azure; CI/CD with Azure DevOps; monitoring and logging.
• Produce comprehensive technical documentation, runbooks, and governance artifacts.
• Conduct knowledge transfer sessions with INT teams.

Skills

Python

LLMs

RAG

API development

DevOps

Education

Bachelor's degree in Computer Science, AI/ML, Data Science, Software Engineering

Tools

LangGraph

AutoGen

Semantic Kernel

CrewAI

LlamaIndex

Job description

The World Bank Group is seeking an ET Consultant 1 – AI Engineer to design, develop, and deploy AI agent systems that integrate with Azure, Databricks, and internal APIs. You will build LLM-enabled features, orchestrate multi-step workflows, and implement responsible AI guardrails.

Strong collaboration across cross-functional teams is essential. You will also produce architecture docs, runbooks, and training materials while ensuring security, data governance, and scalable design in a regulated,