fandishemussa/ResearchHub-Ethiopia

ResearchHub Ethiopia is a scalable, AI-powered platform for aggregating institutional research repositories, managing scholarly metadata, enabling semantic discovery, and supporting research analytics across Ethiopian universities. # ResearchHub Ethiopia > Full-text research intelligence and semantic retrieval are operational workflows, not UI-only features. See the implementation audit, AI intelligence, and semantic search for decision flows, security controls, repair procedures, and current evidence limits. > Local AI operators: see Ollama context management for the 16 GB laptop profile, dynamic RAG budgeting, memory guard, and benchmark commands. ResearchHub Ethiopia is an AI-assisted research information management platform for aggregating, normalizing, searching, and analysing scholarly output from Ethiopian universities. It connects institutional repositories through OAI-PMH, DSpace REST Discovery, or validated file imports, preserves metadata provenance, supports semantic discovery with PostgreSQL and pgvector, and provides operational workflows for harvesting, import review, research intelligence, and repository health monitoring. > **Project status:** active development. The architecture includes scalability and observability foundations. ## Start here - **Run the platform:** Quick start with Docker - **Connect a university repository:** Adding a research source - **Import an existing dataset:** Importing metadata files - **Index downloaded PDFs:** Full-text document indexing - **Configure local AI:** AI providers and Ollama - **Develop and test:** Development without Docker and Testing and quality checks - **Operate and diagnose:** Health and observability, Backup and recovery, and Troubleshooting - **Understand delivery status:** Feature maturity and Known limitations ## Feature maturity The table uses the evidence rules in the enterprise prototype audit. A route, model, or screen alone is not treated as fully verified, and runtime-dependent claims remain pending until exercised on the target stack. | Area | Status | Current capability …