The four things that make a verified resource graph defensible.
Tiered verification, Resource Confidence Score, HSDS v3.0 native, continuous verification cadence. The same four anchors throughout the site, in more depth.
Stricter validation where the consequences of being wrong are highest.
Every facility is classified by the urgency of the most critical service it offers. The classification determines how strictly each field on each record gets validated. A wrong phone number at a crisis center has different consequences than a wrong description on a support group listing. CommonLight reflects that in how the data is verified.
Tier classification feeds a validation matrix that matches each field to a verification method appropriate to what is at stake. The most consequential fields on the most urgent records receive the most rigorous verification. The least consequential fields on routine records receive automated approval.
Every record carries a score. Programs decide the threshold.
The Resource Confidence Score combines four dimensions into one per-record number: completeness, validity, freshness, and provenance. The score is exposed in the API and sortable at query time.
Closed-loop programs can require a minimum confidence score for any record they act on. Records that fall below threshold are flagged, not deleted. Programs see what is borderline, what is fresh, and what needs re-verification. The score is not a marketing number. It is a diligence input.
Are the fields a program needs actually populated?
Do the values pass structural checks against the schema?
How recently was the record verified?
What source, what method, and at what tier?
Built on the standard, not retrofitted to it.
HSDS v3.0 is the open data standard for community resource information. CommonLight is built on HSDS v3.0 from the data layer up, with the Org-to-Facility-to-Service hierarchy preserved end to end. Records flow into existing referral pipes that speak HSDS without translation loss.
A single agency with five clinics and twenty programs is represented as one organization, five facilities, and twenty services — every time.
Org, Facility, and Service relationships preserved across the graph.
HSDS v3.0 across nine endpoints.
Payloads flow into HSDS-aware referral platforms without translation.
Verified on a schedule, not scraped once and stored.
The data is verified on a defined cadence rather than ingested once and assumed permanent. Records re-enter the verification pipeline based on tier classification, freshness rules, and partner-specific configuration. Diff reports surface what changed, when, and against which source.
Partners receive scoped diffs for the geographies and service categories they care about, on the cadence the SOW defines. The graph is a living record of what is true now, not a snapshot of what was true at ingestion.
Where to go from here.
Network Integrity Report.
A scoped audit of your existing resource directory against HSDS v3.0 and our verification rubric. We tell you what is sound, what is stale, and what will not survive closed-loop reporting.
Request a report →EngineeringAPI access.
Schema reference, endpoint documentation, and integration patterns for technical reviewers evaluating fit.
Request access →Fit conversationTalk to our team.
Thirty-minute fit conversation. We will tell you whether your use case is something we can support today, on the roadmap, or out of scope.
Schedule a call →
The verified resource layer underneath closed-loop referral programs.