A web-based decision-support platform — not a search tool, not an AI writer.
Aurora is organized around the actual shape of grant review work. Every surface exists because a real workflow needed it.
Opportunity coverage across major donor ecosystems
Nine working surfaces, mapped to real review steps.
Dashboard
High-signal summary of what deserves attention this week — urgent deadlines, fresh calls, and decisions in progress.
Opportunities list
Full inventory with search, filters, and saved views. Every row is scored against your institutional profile.
Opportunity detail
Score breakdown with rationale per criterion, major risks, missing info, and a suggested next action.
Institutional profile
Thematic priorities, geographies, capacity areas, partner preferences, and funding comfort — editable anytime.
Matching rules
Editable criterion weights with transparent impact on recommendations. No hidden calibration.
Reviews & notes
Reviewer attribution, status transitions, and override history per opportunity.
Alerts
Rules-based alerts — by score, theme, deadline proximity, or geography.
Analytics
Pipeline health, review throughput, and score distribution over time.
Data sources
Visibility and control over every source — portal, newsletter, RSS, manual, or API.
Every opportunity, 14 structured fields.
The schema is designed so data can later come from spreadsheets, APIs, or automated ingestion — without changing a single UI contract.
- Autonomous proposal writing
- Full LLM reasoning layer as primary logic
- Global public grant marketplace
- Real-time sync across all donors (yet)
- Enterprise-grade RBAC (yet)
These are not missing features. They are intentional choices to protect trust before automation.
What we'll measure during a pilot.
- Opportunities reviewed per week per reviewer
- Time-to-triage for a new opportunity
- Percentage of reviewed items with structured rationale
- Override rate (signals trust in scoring)
- Conversion from review → internal application decision
- Pilot retention and continued usage