The Big Why
Frontline humanitarians lack rapid, practical access to contextualised evidence and advice. Existing evidence repositories are fragmented; helpdesks are agency specific, scarce and often project-bound.
What will change?
Field teams, cluster coordinators, government counterparts, and donors requiring fast, contextualised evidence will make better decisions. Shifts evidence use from supply-driven to demand-led. Speeds up learning loops; elevates local questions; reduces duplication of desk reviews. Over time, builds a searchable body of Q&As that maps what the system actually needs to inform a wider research agenda.
What the idea involves
Product: A web platform with (a) an intake wizard that translates practitioner questions into searchable queries; (b) a triage dashboard that routes tickets to a researcher bench by domain/region/language; (c) a response workspace with AI-assisted literature triage, citation capture, and templated ‘2‑pager’ outputs; (d) a public library of anonymised briefs; (e) feedback and up‑vote features.
Service: 3 service tiers — Rapid (24–72h best‑effort synthesis), Deep Dive (1–3 weeks mini‑review), and Advisory (ongoing accompaniment). All responses include quality grading, context flags, and practice tips.
Community & Governance: Curated researcher network (global North/South), local peer reviewers, ethics and conflict‑of‑interest policy, language coverage, and accessibility standards.
Data & AI: Evidence graph of sources; AI retrieval to surface prior answers; human-in-the-loop red‑teaming for sensitive topics; privacy controls for non-public questions.
What would it take to deliver it?
Delivery process: Hosted service with grants and/or fee-for-service for sustainability; MOUs with networks for supply of experts.
Structure: Start embedded in an existing evidence/innovation organisation; evolve to consortium-based governance.
Scope & timeline: 2-region pilot for 18–24 months; global expansion after independent evaluation.
Staffing needs: Managing editor, product manager, research leads (methodology, qual/quant), ontology engineer, AI engineer, librarian, local reviewers, operations.
Partnerships required: Evidence groups, universities, humanitarian agencies, translation providers, tech vendors.
Key capacities: Evidence synthesis, humanitarian analysis, service design, multilingual operations, data protection.
Where it stands and what's next?
Stage: Concept with precedents.
Next steps: Assess potential user uptake vs knowledge generation within agencies and from other organisations (CALP, ALNAP, etc). Who could be better placed to host this?
Opportunities to get involved
Universities and networks to join the bench; agencies to pilot questions; donors to fund pilot and impact evaluation; technologists to support development.