A complete giving workflow
From a charity name to a confident decision.
cleargiving.io gives your foundation a structured, repeatable workflow: from the moment you encounter a charity to the day you document your funding decision.
Step 01
Everything starts with your workspace.
Before your first assessment, cleargiving.io walks you through configuring your foundation's profile. Your mission statement. Your geographic focus. The causes you prioritise. This isn't boilerplate. It shapes how every charity is assessed.
- Name your foundation and record your mission
- Set your geographic focus and strategic priorities
- Define the cause areas you fund: education, health, arts, environment, and more
Step 02
Your scoring model is the intelligence behind every score.
Most AI tools apply a generic lens. cleargiving.io doesn't. Your scoring model defines the exact criteria your foundation uses to evaluate charities, along with the weight you assign each one. Every assessment runs through that lens, not ours.
You might weight 'community rootedness' at 20% and 'financial transparency' at 30%. Another foundation might flip those entirely. That difference matters. It shapes the score, the rationale, and the questions surfaced.
"The scoring model isn't ours. It's yours. We provide the engine. You provide the judgement."
Step 03
Decisions are better made together.
Invite trustees, analysts, and directors to your workspace. Everyone sees the same evidence, the same scores, the same provenance. No more emailed PDFs or verbal summaries that lose nuance on the way.
- Role-based access: analysts assess, directors review, trustees decide
- Comments and discussion threads per assessment
- Shared decision record with named signatories
Step 04
Tell us the charity. We do the research.
You enter a charity name or registration number. Our evidence engine goes to work, cross-referencing public registries, official accounts, and web sources against your scoring model criteria.
This is not a simple API call to an AI. Our pipeline runs a research-first pass: the model is explicitly instructed to surface what it knows from Charity Commission filings, Companies House records, and the charity's own published material before it scores a single criterion. Evidence is gathered. Sources are recorded. Nothing is assumed.
3 public registries cross-referenced per assessment
4 evidence quality grades: Strong, Limited, For Review, Missing
0 assumed facts. Every gap is surfaced explicitly.
Step 05
The AI recommends. You decide.
Every criterion score comes with a rationale, a confidence rating, cited sources, and an evidence quality grade. You can agree, challenge, or override, and every override requires a documented reason.
The audit trail captures both the AI's position and your judgement, side by side. That record is yours.
Ready to set up your workspace?
Request access and we'll walk you through configuration.