CrowdAfrica
CrowdAfrica is a consent-first, offline-capable oral history ingest and public archive platform. Field researchers capture interviews on mobile, manage narrator consent through every stage, and publish to a searchable public archive — with immutable preservation and automated transcript drafting built in.
Key Features
From field capture to public archive — with consent management, preservation, and search at every step.
Interview Workspace
Pre-interview access planning for elders, gatekeepers, and partner contacts. Voice briefing to capture context before recording. Narrator profiles linking multiple sessions with the same person. Rich metadata: language, historical period, recording location, event location, themes, and people mentioned.
Consent Management
Free and informed consent captured at submission — and withdrawable at any stage thereafter. Consent scope controls what can be published, transcribed, shared with partners, or exported. Withdrawal propagates automatically to the archive and any downstream copies. Consent is a first-class object, not a checkbox.
Preservation Pipeline
Validation worker checks format, duration, and metadata completeness. Preservation worker converts to archival formats (FLAC, PDF/A, TIFF) and writes to immutable object storage. Every artifact — audio, video, document, photo, transcript — follows the same consent and preservation rules as the primary interview.
Transcript & OCR Drafting
Automated first-pass transcription from audio and video via the transcription worker. OCR drafts from scanned documents and photos. Drafts are flagged as machine-generated and held for human review before publication. Speaker diarisation and language detection inform transcript structure.
Public Archive
A searchable, browsable public surface for consented materials. Full-text search via OpenSearch. Geographic browse via PostGIS. Filter by language, historical period, theme, narrator, and collection. OAI-PMH endpoint for harvesting by national archives and academic repositories. Embed-ready media player.
Offline-First Mobile
Flutter mobile app for field capture without connectivity. Interviews, metadata, consent, and artifacts queue locally and sync when connection is restored. Non-interrupting coaching prompts during recording guide the interviewer without disrupting the narrator. Works on Android devices common in field research contexts.
Technical Specifications
Backend
- Fastify API (TypeScript)
- PostgreSQL + PostGIS (geo-spatial)
- Redis + worker queue
- OpenSearch (full-text indexing)
Storage
- S3-compatible (MinIO, AWS S3, Linode, IDrive)
- Signed URLs for secure upload/download
- Archival format conversion (FLAC, PDF/A, TIFF)
- Immutable write — preservation records never mutated
Surfaces
- Web intake (Next.js)
- Admin review & moderation
- Public archive (searchable)
- Flutter mobile (offline-first field app)
Standards
- OAI-PMH (metadata harvesting)
- Dublin Core metadata schema
- Kenya DPA 2019 (consent & erasure)
- OAIS reference model for digital preservation
Preserve what matters
For oral history organisations, cultural institutions, university archives, and community memory projects across Africa. Contact us to discuss your collection and deployment needs.