AI Copilot Agent Workspace
Role-aware, source-grounded copilot and specialised agents. Every response shows confidence, evidence completeness, source references, caveats and human review requirement.
Copilot chat
Tailored to Country Director / Strategy Lead
Synthesise the five-year strategic bets and flag where evidence is still emerging before the board visioning session.
AI agents
Specialised, guard-railed agents
Strategy Synthesis Agent
Five-year visioning → Strategic priorities, bets, risks, learning questions
Labels evidence strength & assumptions
Rights Convergence Agent
FHFI / village work → Entitlement gaps & next government follow-up
No rights-achieved claims without evidence
MEL Quality Agent
Data review → Missing data, outliers, confidence labels
Human MEL approval required
Donor Report Agent
BMZ/EU/SWITCH reporting → Draft donor report & annex
Human approval required
Nutrition Smart Agent
NSV village digital twin → Village nutrition risk summary
Flags sensitive data handling
Partner Risk Agent
Partner management → Partner risk memo & capacity plan
Does not auto-suspend partners
Evidence Pack Agent
Reporting & comms → Source-grounded evidence pack
Consent & source checks required
Storytelling Agent
Campaigns & strategy docs → Strategy narrative & public story draft
Marks synthetic/demo text
Agent library
Specialised, guard-railed agents — click Run agent to open the workspace modal
Needs Assessment Agent
AvailableSynthesizes deprivation patterns and suggests first 90-day interventions.
Sample prompt: Run on high-risk clusters
Theory of Change and Logframe Agent
AvailableTurns needs and evidence into outcomes indicators assumptions and risks.
Sample prompt: Generate ToC draft
KPI Trend and Contribution Agent
AvailableCompares frontline trends with baseline and flags contribution confidence.
Sample prompt: Explain KPI change
Midline Adaptation Agent
AvailableConverts midline signals into adaptive management options.
Sample prompt: Prepare midline review
Map Intelligence Agent
AvailableFinds overlaps between vulnerability evidence gaps and partner coverage.
Sample prompt: Run map hotspot analysis
Communications and Storytelling Agent
AvailableCreates narratives only from approved evidence and consent-reviewed assets.
Sample prompt: Draft evidence-backed story
Donor Report Agent
AvailableBuilds donor narrative from KPI evidence finance risks and adaptation actions.
Sample prompt: Generate BMZ report preview
Baseline Data Quality Agent
AvailableSpecialised guard-railed agent. Output requires human review before external use.
Programme Lifecycle Agent
AvailableSpecialised guard-railed agent. Output requires human review before external use.
Frontline Monitoring Agent
AvailableSpecialised guard-railed agent. Output requires human review before external use.
Endline Evaluation Agent
AvailableSpecialised guard-railed agent. Output requires human review before external use.
Rights Convergence Agent
AvailableSpecialised guard-railed agent. Output requires human review before external use.
Evidence Gap Agent
AvailableSpecialised guard-railed agent. Output requires human review before external use.
Partner Capacity Agent
AvailableSpecialised guard-railed agent. Output requires human review before external use.
Run history
Recent agent runs with outputs and review status