How Cause Compass turns public records into discovery.

Cause Compass reshapes public nonprofit data into something you can wander through. It does not rate, certify, or rank organizations. It exists to make it easier to see what a nonprofit is, where it works, and which causes it connects to. This page documents how that happens, and where it stops.

Where the data comes from

The base directory is IRS Exempt Organizations data. Parsley, the import package in this monorepo, processes those files and expands the codes inside them. A second layer adds public web information from organization sites, used only when the match is reasonably confident.

How records become profiles

Import IRS records
Parsley reads state-level IRS Exempt Organizations CSV files and maps each row into a typed nonprofit record keyed by EIN.
Attach reference dictionaries
IRS codes are expanded through local dictionaries for NTEE, activity codes, deductibility, foundation type, filing requirements, affiliation, assets, organization type, and status.
Filter and transform
Eligible organizations become Convex-ready JSONL records with normalized names, addresses, financial buckets, NTEE major codes, and import metadata.
Find public context
The scraper looks for likely official websites, about pages, donation links, social links, and public text that helps describe what the organization actually does.
Structure for discovery
AI-assisted steps turn that public web content into consistent fields: mission, tagline, summary, activities, audience, geography, and keywords.

Which fields come from where

Each profile is assembled from four kinds of input. This is the split we track internally.

Pulled from IRS data

EIN
Legal name
Street, city, state, and ZIP
In-care-of name
NTEE code and NTEE major
IRS activity codes
Subsection, affiliation, classification, ruling, group exemption, status, organization, deductibility, and foundation codes
Asset amount, income amount, and revenue amount before bucketing

Derived by Cause Compass

Slug, generated from organization name plus EIN
Asset bucket and income bucket, derived from IRS financial amounts
Cause browse bucket, derived from NTEE major code
Enrichment stage, derived from pipeline progress

Found by scraping

Donation links
Social media links
Logo links
Email addresses
About page links
Newsletter signup signal
Crawled text used as AI input

Generated or classified with AI

Likely official website URL
Mission
Tagline
One-sentence summary
Why support
Target audience
Geographic focus / reach
Activities
Keywords

What the fields mean

EIN: The employer identification number. Cause Compass uses it as the stable public identifier for IRS-sourced organization records.
NTEE code: The National Taxonomy of Exempt Entities code. It describes the organization type at a more specific level, such as a particular education, health, arts, or human services category.
NTEE major: The first letter of the NTEE code, stored separately because it is useful for broad browsing and aggregate counts.
Activity codes: IRS activity codes describe activities reported in the EO file. A record can contain up to three.
Deductibility code: An IRS code describing whether contributions to the organization are deductible and under what conditions.
Foundation code: An IRS classification for foundation status, such as private foundation, operating foundation, or public charity.
Asset and income buckets: Broad IRS-derived financial ranges, stored instead of making exact financial size the focus of discovery.
Enrichment stage: An internal status for how much useful public profile information has been found and structured for an organization.

IRS codes and our cause buckets

NTEE codes are hierarchical. We store both the specific code and its major, which is the first letter: a specific Education code rolls up to major B. We then group those IRS majors into broader cause areas that are easier to browse.

Arts & Culture: AArts, Culture & Humanities
Education & Youth: BEducation
OYouth Development
Health & Mental Health: EHealth Care
FMental Health & Crisis Intervention
GVoluntary Health Associations & Medical Disciplines
HMedical Research
Environment & Animals: CEnvironment
DAnimal-Related
Human Services: PHuman Services
Housing & Community: LHousing & Shelter
SCommunity Improvement & Capacity Building
Food & Agriculture: KFood, Agriculture & Nutrition
Civil Rights & Legal Aid: ICrime & Legal-Related
RCivil Rights, Social Action & Advocacy
Work & Economic Opportunity: JEmployment
Public Safety & Disaster Response: MPublic Safety, Disaster Preparedness & Relief
Recreation & Sports: NRecreation & Sports
International Affairs: QInternational, Foreign Affairs & National Security
Science, Tech & Research: UScience & Technology
VSocial Science
Philanthropy & Public Benefit: TPhilanthropy, Voluntarism & Grantmaking Foundations
WPublic & Societal Benefit
Faith-Based & Religious Organizations: XReligion-Related
Membership & Mutual Benefit: YMutual & Membership Benefit
Other: ZUnknown
—No NTEE major code

How we describe organization scale

Exact finances are not the story here. Broad size language conveys rough scale without turning discovery into financial ranking, and geographic reach is inferred from public website content rather than declared.

Size, from IRS financial amounts

Parsley converts IRS asset and income amounts into shared buckets. The app primarily reads the asset bucket.

micro: Under $50,000
small: $50,000 to $249,999
mid: $250,000 to $999,999
large: $1 million to $9,999,999
mega: $10 million or more
unknown: Missing or unavailable

Reach, inferred by AI

Geographic focus is read from public website content and constrained to four labels.

Local: Rooted in a city, neighborhood, county, or nearby community.
Regional: Spans a metro area, multi-county area, state, or region.
National: Operates across the United States or serves a national audience.
Global: Crosses national borders or focuses on international issues.

How size appears in the product

Grassroots: micro, small; Smaller organizations, by IRS asset bucket.
Established: mid; Mid-sized organizations.
Institutional: large, mega; Large and major organizations.

What this methodology does not do

Cause Compass does not verify every claim an organization makes, inspect financial filings for quality, or assign scores. Public records lag reality, organization websites move, and AI-assisted summaries can be incomplete or wrong.

The product is built for discovery. Before donating, volunteering, or making any important decision, review the organization directly and consult the original public sources.

Where the data comes from

How records become profiles

Import IRS records

Parsley reads state-level IRS Exempt Organizations CSV files and maps each row into a typed nonprofit record keyed by EIN.

Attach reference dictionaries

IRS codes are expanded through local dictionaries for NTEE, activity codes, deductibility, foundation type, filing requirements, affiliation, assets, organization type, and status.

Filter and transform

Eligible organizations become Convex-ready JSONL records with normalized names, addresses, financial buckets, NTEE major codes, and import metadata.

Find public context

The scraper looks for likely official websites, about pages, donation links, social links, and public text that helps describe what the organization actually does.

Structure for discovery

AI-assisted steps turn that public web content into consistent fields: mission, tagline, summary, activities, audience, geography, and keywords.

Which fields come from where

Each profile is assembled from four kinds of input. This is the split we track internally.

Pulled from IRS data

EIN
Legal name
Street, city, state, and ZIP
In-care-of name
NTEE code and NTEE major
IRS activity codes
Subsection, affiliation, classification, ruling, group exemption, status, organization, deductibility, and foundation codes
Asset amount, income amount, and revenue amount before bucketing

Derived by Cause Compass

Slug, generated from organization name plus EIN
Asset bucket and income bucket, derived from IRS financial amounts
Cause browse bucket, derived from NTEE major code
Enrichment stage, derived from pipeline progress

Found by scraping

Donation links
Social media links
Logo links
Email addresses
About page links
Newsletter signup signal
Crawled text used as AI input

Generated or classified with AI

Likely official website URL
Mission
Tagline
One-sentence summary
Why support
Target audience
Geographic focus / reach
Activities
Keywords

What the fields mean

EIN

The employer identification number. Cause Compass uses it as the stable public identifier for IRS-sourced organization records.

NTEE code

The National Taxonomy of Exempt Entities code. It describes the organization type at a more specific level, such as a particular education, health, arts, or human services category.

NTEE major

The first letter of the NTEE code, stored separately because it is useful for broad browsing and aggregate counts.

Activity codes

IRS activity codes describe activities reported in the EO file. A record can contain up to three.

Deductibility code

An IRS code describing whether contributions to the organization are deductible and under what conditions.

Foundation code

An IRS classification for foundation status, such as private foundation, operating foundation, or public charity.

Asset and income buckets

Broad IRS-derived financial ranges, stored instead of making exact financial size the focus of discovery.

Enrichment stage

An internal status for how much useful public profile information has been found and structured for an organization.