The AI Applications Developer will design and build an AI-powered, RAG based knowledge retrieval system under the next phase of the Platform for Development Research and Communication (PDRC 2.0).
This engine will systematically convert the six Human Development volumes into high-quality, structured dissemination outputs including policy briefs, teaching case studies, executive summaries, translation-ready drafts, event notes and multimedia scripts.
The role requires good understanding of Artificial Intelligence concepts, development of Large Language Models, and building production ready RAG (Retrieval-Augmented Generation) systems.
The position will play a foundational role in establishing a scalable, AI-enabled knowledge infrastructure for future research outputs.
The Platform for Development Research and Communication (PDRC) has produced six comprehensive Human Development volumes synthesising evidence across nutrition, health, education, institutions, climate and rural prosperity.
The next phase focuses on translating these volumes into accessible, policy-relevant and high-impact outputs at scale. To enable systematic and efficient transformation of large research documents into structured dissemination products, the project seeks to build an AI-driven Knowledge Engine. This system will:
- Extract thematic insights across chapters
- Generate structured policy briefs
- Create case study drafts
- Develop executive summaries
- Prepare translation-ready formats
- Assist in event briefing notes
- Support multimedia scripting
The AI Applications Developer will lead the design, development, testing and deployment of this system.
The successful candidate will carry out the following activities:
- Knowledge Engine Development: Design and develop a web-based, natural language query interface over the 6 PDRC volumes. Build a data pipeline that:
a) Chunks, embeds, and indexes the PDF volumes into a vector store
b) Retrieves relevant passages at query-time
c) Generates grounded answers with source citations via an LLM (OpenAI API or open-source equivalent)
d) Surfaces associated data and charts on demand. The frontend shall be built using React or an equivalent framework; the backend shall expose an API layer connecting the UI to the RAG pipeline
- Data Engineering: Develop data pipelines that enables the system to:
a) Perform structured document parsing
b) Generate chunks and embeddings
c) Metadata generation
d) Thematic clustering
e) Citation preservation systems
Ensure traceability between source content and generated outputs.
- Quality Assurance: Develop test cases to validate the system against:
a) Hallucination detection
b) Fact-grounding checks
c) Bias assessment
- Deployment: Work with other tech personnel in the department to deploy the data pipelines, backend and frontend to the production servers.
- Collaboration with Research & Communications Teams: Work closely with:
a) Researchers (content structuring)
b) Events Team (brief generation tools)
c) Communications & Design Cluster (structured output generation)
