Ten transparent dimensions. Every field cited.
Catylst reasons across ten dimensions of donor-organization fit, cites the profile field behind every claim, and offers three levels of explanation depth for every match.
Every dimension has a donor field and an organization field.
When both exist, we cite them. When one is missing, we say so. No composite score, no paid placement, no hallucination.
Cause
Compares your top cause priorities with the organization's tagged cause areas so overlap is counted, not guessed.
Geography
Matches your preferred giving regions to where the organization actually operates — country, region, and community.
Stage
Surfaces whether the organization is seeding, scaling, or sustaining so your grant lands at the right maturity.
Trust tier
Counts how many of the seven verification badges the organization has earned against your stated minimum.
Budget fit
Compares your typical grant size against the organization's active funding range so the dollars are plausibly useful.
Faith alignmentOptional
Opt-in match of your faith identity and ministry orientation with the organization's faith profile. Skipping has zero penalty.
Workforce
Reveals the people doing the work — staff, volunteers, local partners — so you know who your grant supports.
Maturity
Weighs governance, systems, and reporting discipline so you can fund organizations ready to steward the capital.
Risk profile
Surfaces operational, regulatory, and reputational exposures honestly, flagging gaps rather than concealing them.
Reporting cadence
Matches your expected follow-up rhythm with the organization's actual reporting schedule — six-month, annual, or custom.
Read the same match at three zoom levels.
Toggle between compact, standard, and detailed. Every density cites the same profile fields; only the prose depth changes.
Standard:Strong fit across cause, geography, and trust tier. Budget fit is within one bracket. Faith alignment is opted in and matched. Reporting cadence is annual versus your stated six-month preference.
The no-fabrication guarantee.
Every match and every grant package passes through a five-layer validation pipeline. Each layer exists to prevent a specific category of fabrication — together they are the reason every claim on Catylst is cited rather than invented.
Structural sentinels
Required fields and citation anchors are asserted before any AI output is emitted, so missing data fails loudly.
Cite-bound conjunction
Every claim is bound to a profile-field citation; unbound sentences are rejected at generation time, not at review.
Symmetric leak check
Donor-profile facts cannot leak into organization-profile output and vice versa, enforced symmetrically across both sides.
CI lint gateway
Every pull request runs structural, citation, and leak linters in CI. Builds that would ship fabricated output fail before merge.
Output wrapper lint
The rendering wrapper re-runs the full lint suite on the final payload so production cannot emit content the linter would reject.