Engagement
78 / 100
↑ above baseline, cadence fading
- _BASELINE
- 66
- _DELTA
- +12
- _CONF
- 0.84
Sample analysis · 33 fields
One student, end-to-end.
The numbers are fake. The shapes are real. This is exactly what your Slate Query Service
sends Celia, what Celia reasons on, and what gets written back into the student’s
record in 33 filterable ss_celia_* fields.
Tenant org_ncsu · Flow Summer Melt 2026 — CS Cohort · Model celia-v1.4
1. What Celia saw (from your Slate Query Service)
Anonymized. Behavioral. No names, no emails, no phones, no addresses, no financial account numbers. That is not a policy — it is a structural property of the query itself.
{
"slate_person_ref": "P-142098",
"slate_person_guid": "8f42a91c-0000-4000-8000-000000000000",
"flow_name": "Summer Melt 2026 — CS Cohort",
"stage": "app_submitted",
"days_in_stage": 23,
"docs_missing": ["fafsa_submitted", "counselor_recommendation_letter"],
"last_touch": { "channel": "email", "days_ago": 7, "opened": false },
"engagement_7d": { "logins": 3, "portal_clicks": 15,
"deepest_page": "/financial-aid/packages" },
"cohort_base": { "program_yield_rate": 0.68, "stage_median_days": 9 }
}PII fields never sent
name · email · phone · address · DOB · SSN · essay text · recommendation letter text
Architectural, not aspirational
Your CIO can verify this at the architecture level — it’s the shape of the query, not a promise we make.
2. What Celia wrote back into Slate
33 structured fields, every one of them filterable, reportable, and list-buildable in Slate. Organized in four Layers — identifiers, the triage list-view, the drill-down, and metadata + counselor feedback. Your team works in Slate; Celia just makes Slate smarter.
Layer 0
Identifiers — 3 fields (join key + Flow context)
slate_person_guid
8f42a91c-0000-4000-8000-000000000000
slate_person_ref
P-142098
ss_celia_flow_name
Summer Melt 2026 — CS Cohort
Layer 1
List view — 5 fields (morning triage)
ss_celia_risk
High
ss_celia_engagement
78
ss_celia_readiness
42
ss_celia_yield
0.61
ss_celia_next_action_1
Send FAFSA reminder today
Layer 2
Student record drill-down — 15 fields
Readiness
42 / 100
↓ stalled 23 days, 2 blocking items
- _BASELINE
- 65
- _DELTA
- −23
- _CONF
- 0.91
Yield
0.61 prob
high-likely; readiness drag −0.14
- _BASELINE
- 0.68
- _DELTA
- −0.07
- _CONF
- 0.78
ss_celia_risk_confidence + _context
0.82 confidence
Historical melt rate for this stage + timeline pattern: 38% if not touched within the next 7 days.
ss_celia_next_action_2
Personal counselor call within 72h — walk through financial-aid timeline
Layer 3
Metadata + feedback — 10 fields
COMPUTED_AT
2026-04-23 02:14 UTC
DATA_AS_OF
2026-04-22 23:59 UTC
NEXT_REFRESH
2026-04-24 02:00 UTC
MODEL_VERSION
celia-v1.4
ss_celia_next_action_3 + deadline
Nudge recommender on letter — offer auto-fill submission link
Deadline: 2026-04-26
ss_celia_next_action_why
FAFSA is 14d overdue; cohort stage-median is 9d. Recommendation letter blocks further progress. Yield probability will drop another ˜10% if both gaps remain unclosed in 7d.
ss_celia_signals_explained
Engagement 78 driven by 3 logins in 7 days (baseline 1.2) and deep financial-aid page browsing; offset by a 7-day unopened email. Readiness 42 pulled down by FAFSA and recommendation letter both overdue. Yield 0.61 high-likely but with readiness drag.
ss_celia_acknowledged_by
null (pending counselor action)
ss_celia_ack_outcome
pending
3. What your counselor opens in Slate
No new tool to learn. No dashboard to babysit. The next morning, the counselor opens the student record in Slate and sees this at the top:
P-142098 · Computer Science · Fall 2026
Jordan R. · s_8f42a91c
Flow: Summer Melt 2026 — CS Cohort
Engagement
78
↑ +12 vs cohort 66
Readiness
42
↓ −23 vs cohort 65
Yield probability
0.61
↓ −0.07 vs 0.68
Do these today (deadline 2026-04-26)
Send FAFSA reminder · Call within 72h · Nudge recommender
Less than 3 seconds from opening Slate to knowing what to do.
4. The economics, per student
Every Celia run is logged with per-tenant cost attribution. This is what one student cost to analyze on the batch this sample came from.
Input tokens
1,420
Output tokens
380
Batch size
20
Cost / student
$0.00021
Scoring an entire 10,000-student cohort every night runs around $2.10/day.
90-day free pilot
Pilot Celia on your Slate. Free for 90 days.
Tell us about your institution and your Slate setup. We'll book your kickoff, provision your workspace on the call, and you have 90 days to verify Celia moves the metric you care about. No card. No commitment. Real Slate integration.