Definition
Alumni data enrichment is record improvement, not record creation.
The process begins with an alumni record the school already recognizes. Enrichment attempts to append or update useful professional facts—such as employer, role, college, location, and LinkedIn profile—without changing the underlying identity of that record.
That distinction matters. A school information system can tell Studiously that a person is an alumnus and provide identifying context. It is not automatically a reliable source for that person's current career. Career information is matched from separate professional sources and must clear a different set of checks.
Veracross boundary: Studiously can pull alumni information from Veracross. Veracross does not provide the career information used in this enrichment process, so employer, title, and career history must come from separate professional-data sources. Studiously does not write enriched data back to Veracross through the API.
Inputs
What enters the matching process
Matching quality depends on the identifying context available for each person. Studiously always begins with the name on the school's roster. School, class year, college, and other identifiers can narrow the candidate set when they are available and selected for the run.
| Source | What it may contribute | How it is used |
|---|---|---|
| School roster | Name, high school, class year, email, and location when available | Establishes the alumni record and its school |
| Veracross | Alumni identity and roster information available through the school's connection | Imports alumni; it is not the source of current career data |
| Raiser's Edge NXT | Constituent, education, contact, location, and existing business context | Seeds the roster and provides comparison signals |
| CSV | The mapped fields included in the school's source export | Supports schools without a direct source connection and preserves the original baseline |
| Professional-data source | Possible employment, education, location, profile URL, skills, and contact information | Supplies the candidate professional record that must be validated |
Depending on the school's configuration and the fields already present, a match request may also include an existing email, LinkedIn URL, employer, job title, location, phone number, or birth information. Those optional signals are used to disambiguate people; they are not assumed to be current merely because the school already has them.
Process
From roster row to reviewable profile
-
Import and school assignment
The source record is normalized and attached to the school it came from before any enrichment begins.
-
Candidate matching
A professional-data provider evaluates the configured identity signals and may return a candidate profile with a provider likelihood score. A score is evidence, not proof.
-
Raw-result preservation
The provider response, match outcome, settings, and identifiers sent for that attempt are retained for review and audit before live alumni fields are changed.
-
AI-assisted normalization
An extraction step reads the returned employment and education history together, resolves contradictory fields, avoids obvious side jobs and placeholders, and proposes a consistent set of fields. It is instructed not to invent missing information.
-
Validation and deterministic checks
The proposed match is graded for confidence and flags. Separate code checks identifiers and education timelines so high-confidence model output cannot override an impossible chronology or a conflicting profile URL.
-
Review and application
Clean records can follow a tightly gated apply path when enabled. Uncertain records remain reviewable. An operator can accept, edit, or reject the proposed match before it becomes live data.
-
Classification and delivery
Accepted facts can then power industry, seniority, company-type, location, summary, and search fields before selected data is delivered to the directory, RE NXT, or a CSV workflow.
Trust layer
Confidence comes from multiple gates
Studiously does not reduce identity to a single score. Provider likelihood, normalized fields, source comparisons, deterministic checks, and review status all have separate jobs.
Provider likelihood
Measures how strongly the professional-data provider matched the submitted signals. It is not treated as a probability or identity guarantee.
Identifier agreement
An existing email or LinkedIn URL can confirm the candidate when it agrees with the returned record.
Source cross-checks
Education, business, email, and location context from RE NXT can strengthen or weaken the proposed match without being copied blindly.
Wrong-person backstops
Name conflicts, different profile URLs, impossible education dates, and missing critical fields can stop automatic application.
Automatic does not mean unchecked. The strict auto-apply path requires a validated extraction and identity confirmation, is available only when enabled for that school, and blocks additional warning conditions. CRM write-back applies another safety gate before data leaves Studiously.
Human review
What a reviewer can see and decide
Review is designed around comparison, not blind approval. The operator can inspect the school's existing values, the provider's returned record, the normalized proposal, the match score, and record-specific warning messages in one workflow.
- Accept: apply the proposed fields that should become part of the alumni record.
- Edit: choose or correct individual values before committing them.
- Reject: mark the candidate as the wrong person and clear live enrichment fields while retaining the audit history.
- Re-run: try a deliberate new match configuration when better identifying information becomes available.
Accept and reject decisions are recorded separately from provider attempts. That preserves both sides of the history: what the source returned and what a human decided to do with it.
Outputs
What an enriched profile may contain
Output coverage varies by person. A field appearing below means the pipeline can represent it—not that every professional source will return it or that every proposed value will be approved.
Career
Current employer, job title, industry, seniority, company size, skills, source verification date, and a cleaned career history.
Education
Undergraduate college, major, degree, college graduation year, graduate school, graduate degree, and education history.
Location and profile
Current city or region, LinkedIn profile URL, and structured location tags for search and segmentation.
Restricted contact fields
Professional email, personal emails, or phone numbers may be returned. These are handled separately from member-facing professional profile data.
Search classifications
Industry families, company types, seniority, summaries, and search representations derived only after usable profile facts exist.
Audit context
Attempt settings, source outcome, validation flags, review decision, timestamps, and CRM sync history.
Privacy and isolation
The school boundary is set before enrichment
Every alumni record is associated with a school. Enrichment reads and writes use that school association as a filter; a professional-data result or an AI-generated field cannot choose another tenant or expand what data is reachable.
- Operator-restricted raw data: raw provider responses and enrichment review tools are not part of the standard member-facing directory experience.
- Sensitive fields separated: work email, personal email arrays, phone arrays, and the raw provider payload are excluded from standard authenticated alumni-table reads.
- Limited model context: normalization is performed for one alumni record at a time, with only the relevant school name and available CRM cross-reference supplied for that record.
- Untrusted model output: AI may propose field values, but it never controls school access, record ownership, or the destination tenant.
- Safe test environments: development and sandbox controls prevent test activity from silently calling live professional-data, CRM, or AI providers.
Freshness
Refreshes are deliberate, not a vague promise of “always current.”
Studiously records when enrichment ran and preserves a source-provided verification date when one exists. Neither timestamp guarantees that the person updated every public professional source on the same day.
Eligible professional profiles can be included in an operator-run refresh using a chosen staleness window. The refresh compares the new source profile with the prior baseline, detects identity drift or missing profiles, and stages changed employer, title, or location values for review. Clean refreshes may follow an enabled controlled-commit path; uncertain changes stay reviewable.
No universal refresh interval is promised. Timing depends on the school's workflow, source availability, whether a stable professional profile is available, and whether a proposed change clears review. A record can become stale between refresh runs.
Delivery
Where accepted enrichment goes
The destination matters because systems support different read and write capabilities. Studiously does not describe every integration as “two-way” when the underlying API is read-only.
| Destination | Inbound | Enriched-data delivery |
|---|---|---|
| Studiously | Imported roster becomes the school-scoped alumni base | Accepted professional fields power the directory, filters, analytics, and search |
| Raiser's Edge NXT | Constituents and available CRM context can be imported through the Blackbaud connection | Supported enrichment fields can be written through the API to dedicated Constituent custom fields, subject to permissions and safety gates |
| Veracross | Studiously can pull alumni information; Veracross is not the source of career history | No API write-back. Selected profile updates or enriched rows can be exported as CSV for a manual Veracross workflow |
| CSV workflow | A school's mapped source file can seed the alumni roster | Enriched columns can be joined back to the source file by a stable identifier or email when available |
For field-level details on the direct Blackbaud path, see the Raiser's Edge NXT integration guide. For the broader product experience, visit Alumni Intelligence.
Limitations
What enrichment cannot guarantee
- Coverage is uneven. Some alumni have rich professional histories; others have no usable professional record.
- A no-match is not a conclusion. It means the selected signals did not produce a usable result from that source at that time.
- Source data can be stale or contradictory. Current employer, title, and location are especially time-sensitive.
- Common names remain difficult. Sparse identifiers, name changes, and ambiguous education histories increase wrong-person risk.
- Profile URLs are not perfect identifiers. Aliases, redirects, deleted profiles, and duplicate URLs can require manual resolution.
- AI normalization can be wrong. It improves consistency but does not replace source evidence, deterministic checks, or human judgment.
- Veracross does not supply the career layer. Its alumni records help establish the roster; separate professional sources provide possible career information.
- CRM delivery has prerequisites. Direct write-back depends on a connected integration, destination permissions, supported fields, and an accepted match.
The practical standard: enrichment should make an alumni database more useful while making uncertainty visible. A blank or review-needed field is better than a confident-looking wrong person.