Setup & integrations
Slate starter kit
Pre-built Slate artifacts your admin drops into your Slate instance so you don't hand-build them. One Complete Suitcase wires up everything; individual GUIDs are listed for piecemeal imports.
- Who uses this
- Every authenticated team member. Driven by your Slate captain.
What it is
The handoff page from CeliaConnect to your Slate captain. Each row is one suitcase artifact — a bundle of fields, queries, source formats, or dashboards. Click Copy GUID next to a row, paste into Slate's suitcase importer, and the imported item lands in your Slate instance ready to wire up to a Flow.
The first row is a one-click Complete Suitcase that imports everything in one go. The five rows below it are the individual artifacts inside that bundle, listed separately so you can re-import or replace one piece at a time without disturbing the others.
The artifacts
Complete Suitcase
One-click bundle that stands up everything CeliaConnect needs inside Slate — every ss_celia_* custom field (Person + Entity scope), the reference Claude prompt templates,
both sample queries, both writeback source formats, and the Person + Entity dashboards. Import this
once and the five artifacts below are already wired up; the individual GUIDs are listed only if
you prefer to import piecemeal.
Slate Query — In Progress
Reference Slate query Celia reads from on every scoring run. Returns the per-student feature vector — anonymous IDs and dictionary-mapped fields only, no names, emails, phones, or addresses — filtered to students still active in the funnel (no decision yet). Use this query URL as the In Progress query when you create a Flow.
Slate Query — Positive Outcomes
Reference Slate query for labeled training data — last cycle's enrollees and known conversions. Same column shape as the In Progress query, but filtered to confirmed positive outcomes. Celia scores these rows once so the model learns what "good" looks like at your institution; rows here are never written back to.
Source Format (Person)
Person-scope writeback. Celia upserts the latest scoring run onto each Slate person record — engagement, readiness, and yield scores, the narrative summary, and a scored-at timestamp. Last value only: every run overwrites the prior one. Use this when counselors want a single live snapshot in the person view they already work in.
Source Format (Entity)
Entity-scope writeback. Celia appends a new entity row per scoring run, preserving the full history of every score and narrative over the cycle. Use this when you want trend lines, longitudinal mailings, or an audit trail of how a student's readiness evolved week over week.
Dashboard
Reference Slate dashboard layout that surfaces a student's latest CeliaConnect scores (read from the Person source format) alongside the historical score timeline (read from the Entity source format). Drops into your existing Slate person record so your team sees CeliaConnect signal in the same view they already use to work the funnel.
How to import a suitcase
- In Slate, open Database → Search and click Suitcase Import.
- Paste the suitcase GUID from this page.
- Preview the object Slate is about to create and confirm the import.
- Grant the CeliaConnect service user read (queries) or write (source formats, fields) access on the imported item.
Importing never overwrites existing items — Slate creates a new copy you can rename. The sample query and prompts reflect CeliaConnect's reference contract; you can rename, re-scope, or rewrite anything after import — Celia only needs the column aliases to match.
Common workflows
Stand up CeliaConnect in one click
- On this page, click Copy GUID next to Complete Suitcase.
- Send it to your Slate captain with the note "import this as a suitcase".
- Verify the five canonical artifacts landed using the Slate setup kit page.
- When all rows turn green, create your first Flow.
Re-import a single artifact
- Find the row for the artifact you need (queries, source formats, or dashboard).
- Copy its GUID and import as a suitcase in Slate.
- Slate creates a fresh copy alongside the existing one — rename or replace as needed.
Choose Person vs Entity writeback
- Person if your team works in the Slate person view and only needs the latest score.
- Entity if you want trend lines, weekly snapshots, or an audit trail.
- Both if you want today's score in the person view and the historical trail on the entity. The Complete Suitcase imports both — pick which one each Flow targets when you create it.
Common gotchas
Suitcase import fails in Slate
Slate suitcase imports require admin permission and a clean target environment. Try in your sandbox first; once it lands cleanly there, repeat in production.
Dashboard widgets show "no data"
The dashboard reads from the Person and Entity source formats. If neither writeback has run yet (no Flow created, or the Flow is paused), the widgets stay empty. Run a Flow once and refresh the dashboard.
ss_celia_* field roster
The full list of ss_celia_* custom fields the Celia Fields and Prompts suitcase installs
in your Slate. Field names are locked — Celia's writeback expects these exact names to match. Slate
labels and sample values are suggestions; rename or restyle inside Slate freely as long as the field
name stays the same.
Credentials. In Slate, name the user that powers the inbound Query (Username with Web Service Permissions) celiaconnect, and name the service account that authenticates the Source
Format target celiaconnect as well. Both share the same password — one credential
pair, two Slate-side wirings. Paste that single username + password into Settings → Slate connection.
1. Celia fields to create in Slate (31)
| Field name | Data type | Suggested Slate label | Sample value | What this means |
|---|---|---|---|---|
ss_celia_flow_name | Text (80) | Celia · Flow Name | Summer Melt 2026 — CS Cohort | Which Celia Flow produced this row. Lets a counselor tell at a glance which lens is talking when the same person is scored by more than one Flow. |
ss_celia_risk | Text (enum) | Celia · Risk | None · Low · Medium · High · Critical | Celia’s overall risk verdict for this person. Drives the urgency of the next action. |
ss_celia_engagement | Number (integer 0–100) | Celia · Engagement | 78 | Engagement score 0–100. How active this person is right now (logins, opens, replies) relative to their cohort. Higher = more engaged. |
ss_celia_readiness | Number (integer 0–100) | Celia · Readiness | 42 | Readiness score 0–100. How close this person is to being able to make a deposit decision (FAFSA in, transcripts received, blockers cleared). Higher = more ready. |
ss_celia_yield | Number (decimal 0–1) | Celia · Yield | 0.61 | Yield score 0–1. Celia’s estimated likelihood that this person will actually enroll if admitted. Higher = more likely to yield. |
ss_celia_next_action_1 | Text (140) | Celia · Next Action 1 | Send FAFSA reminder today (SMS + email) | The single highest-leverage action a counselor should take next for this person. Short, specific, and verb-first. |
ss_celia_engagement_baseline | Number (integer 0–100) | Celia · Engagement Baseline | 66 | The cohort baseline engagement score (0–100) Celia is comparing this person against — their peer median, derived from the institution’s own positive-outcome history. |
ss_celia_engagement_delta | Number (integer −100–100) | Celia · Engagement Δ | 12 | How far above or below baseline this person’s engagement is, on the same 0–100 scale. Positive = better than peers; negative = behind. |
ss_celia_engagement_conf | Number (decimal 0–1) | Celia · Engagement Conf | 0.84 | How confident Celia is in the engagement score (0–1). Low confidence usually means thin signal — give the score less weight. |
ss_celia_engagement_text | Text (140) | Celia · Engagement Note | ↑ above baseline, cadence fading | One-line counselor-facing note explaining the engagement score. Safe to surface in mailings or counselor dashboards. |
ss_celia_readiness_baseline | Number (integer 0–100) | Celia · Readiness Baseline | 65 | The cohort baseline readiness score (0–100) Celia is comparing this person against, derived from the institution’s own positive-outcome history. |
ss_celia_readiness_delta | Number (integer −100–100) | Celia · Readiness Δ | -23 | How far above or below baseline this person’s readiness is, on the same 0–100 scale. Negative values flag a stall. |
ss_celia_readiness_conf | Number (decimal 0–1) | Celia · Readiness Conf | 0.91 | How confident Celia is in the readiness score (0–1). Pair with the score itself before triaging. |
ss_celia_readiness_text | Text (140) | Celia · Readiness Note | ↓ stalled 23 days, 2 blocking items | One-line counselor-facing note explaining the readiness score. Safe to surface in mailings or counselor dashboards. |
ss_celia_yield_baseline | Number (decimal 0–1) | Celia · Yield Baseline | 0.68 | The cohort baseline yield probability (0–1) Celia is comparing this person against, derived from the institution’s own positive-outcome history. |
ss_celia_yield_delta | Number (decimal −1–1) | Celia · Yield Δ | -0.07 | How far above or below baseline this person’s yield probability is, on the same 0–1 scale. Positive = more likely to yield than peers; negative = less likely. |
ss_celia_yield_conf | Number (decimal 0–1) | Celia · Yield Conf | 0.78 | How confident Celia is in the yield score (0–1). |
ss_celia_yield_text | Text (140) | Celia · Yield Note | high-likely; readiness drag -0.14 | One-line counselor-facing note explaining the yield score. Safe to surface in mailings or counselor dashboards. |
ss_celia_risk_confidence | Number (decimal 0–1) | Celia · Risk Confidence | 0.82 | How confident Celia is in the overall risk verdict (0–1). Pair this with ss_celia_risk before triaging. |
ss_celia_risk_context | Long text (500) | Celia · Risk Context | Historical melt rate for this stage+timeline pattern: 38% if not touched within 7 days. | Free-text explanation of what’s driving the risk verdict — historical pattern, comparable cohort, what tends to go wrong here. Safe to surface to counselors. |
ss_celia_next_action_2 | Text (140) | Celia · Next Action 2 | Personal counselor call within 72h | Second-priority counselor action — what to do after the top-priority one. |
ss_celia_computed_at | Date | Celia · Computed At | 2026-04-23 | The date this scoring run was computed. The “as-of” stamp for everything else in the row. |
ss_celia_data_as_of | Date/Time | Celia · Data As Of | 2026-04-22T23:59:59Z | The cutoff timestamp of the underlying Slate data Celia scored against. Useful when investigating why a score lags reality. |
ss_celia_next_refresh | Date/Time | Celia · Next Refresh | 2026-04-24T02:00:00Z | When the next scoring run is scheduled. Tells a counselor how stale the row is allowed to get. |
ss_celia_model_version | Text (32) | Celia · Model Version | celia-v1.4 | Which Celia model version produced this row. Audit-trail breadcrumb when comparing scores across runs. |
ss_celia_next_action_3 | Text (140) | Celia · Next Action 3 | Nudge recommender on letter | Third-priority counselor action — the next thing after actions 1 and 2. |
ss_celia_next_action_why | Long text (500) | Celia · Next Action Why | FAFSA 14d overdue; cohort median 9d. | Why Celia chose the top action — the one-sentence reasoning a counselor can paste into a personal note. |
ss_celia_next_action_deadline | Date | Celia · Next Action Deadline | 2026-04-26 | The recommended date by which the top action should be taken. Past this date, the action loses leverage. |
ss_celia_signals_explained | Long text (800) | Celia · Signals Explained | Engagement 78 driven by 3 logins/7d. | The longest counselor-facing narrative — what signals are driving the engagement, readiness, and yield scores in plain English. |
ss_celia_acknowledged_by | Text (64), nullable | Celia · Acknowledged By | user_27 | The user ID of the counselor who acknowledged the latest scoring row in CeliaConnect. Read by Celia on the next run; never written by Celia. |
ss_celia_ack_outcome | Text (enum) | Celia · Ack Outcome | pending · completed · not_applicable · escalated | What the counselor did about the latest row. Read by Celia to learn from feedback; never written by Celia. |
2. Sample In Progress query — fields
The In Progress query feeds Celia per-student feature vectors during the active cycle (everyone
who hasn't deposited or withdrawn). Same shape as the Positive Outcomes query — only the WHERE clause
differs:
WHERE stage NOT IN ('deposited','enrolled','withdrawn').
| Field name | Data type | Slate label | Sample value | What this means |
|---|---|---|---|---|
slate_person_ref | Text | — | P-308422 | The institution’s own person reference (e.g. P-308422) — read on the inbound query so Celia can stitch results back to the Slate record. Never a name, email, or phone. |
slate_person_guid | Text (UUID) | — | 9c4d2e7a-0000-4000-8000-000000000017 | Slate’s internal GUID for this person — the match key Celia uses to write the analysis back to the right record. Anonymous identifier only; never PII. |
program | Text | Program | cs_bs | Which academic program this person is in the funnel for. Lets Celia compare them to peers in the same program. |
cohort | Text | Cohort | fall_2026 | The intake cycle this person belongs to. Lets Celia compare them to peers in the same cycle. |
stage | Text (enum) | Stage | inquiry · app_started · app_submitted · admitted · deposited | Where this person is in the funnel today. Drives expectations — an admitted student that hasn’t deposited is a different problem than an inquiry that hasn’t applied. |
stage_entered_at | Date | Stage Entered At | 2026-04-08 | The date this person entered their current stage. Drives the “how long have they been stuck here?” math. |
days_in_stage | Number (integer) | Days in Stage | 18 | How many days this person has sat in the current stage. Higher = more risk of stalling. |
last_touch_days_ago | Number (integer) | Last Touch (days ago) | 11 | Days since the institution last reached out (email, call, SMS). Higher = quieter cadence than the cohort. |
last_login_days_ago | Number (integer) | Last Login (days ago) | 6 | Days since this person last opened a portal session. A leading indicator of fading interest. |
is_positive_outcome | Number (0/1) | Is Positive Outcome | 0 | 1 if this person already converted (deposited or enrolled), 0 if still in the funnel. Drives which rows train the model vs. which rows get scored. |
transitioned_at | Date, nullable | Transitioned At | null | The date this person crossed into a positive outcome (or null if not yet). Used to time-bound the training window. |
ss_celia_acknowledged_by | Text (64), nullable | Celia · Acknowledged By | user_27 | The user ID of the counselor who acknowledged the prior scoring row in CeliaConnect. Read back into Celia on the next run so the model learns from feedback. |
ss_celia_ack_outcome | Text (enum) | Celia · Ack Outcome | pending · completed · not_applicable · escalated | What the counselor did about the prior row. Read back into Celia on the next run so the model learns from feedback. |
ss_celia_risk | Text (enum) | Celia · Risk | None · Low · Medium · High · Critical | The prior run’s risk verdict, read back so Celia can compare run-over-run drift. |
ss_celia_engagement | Number (integer 0–100) | Celia · Engagement | 74 | The prior run’s engagement score (0–100), read back so Celia can compare run-over-run drift. |
ss_celia_readiness | Number (integer 0–100) | Celia · Readiness | 69 | The prior run’s readiness score (0–100), read back so Celia can compare run-over-run drift. |
ss_celia_yield | Number (decimal 0–1) | Celia · Yield | 0.71 | The prior run’s yield score (0–1), read back so Celia can compare run-over-run drift. |
ss_celia_next_action_1 | Text (140) | Celia · Next Action 1 | Confirm deposit by 2026-06-15 | The prior run’s top-priority action, read back so Celia knows what was already recommended. |
ss_celia_next_action_deadline | Date | Celia · Next Action Deadline | 2026-06-15 | The prior run’s deadline for the top action, read back so Celia knows whether the window has passed. |
ss_celia_risk_confidence | Number (decimal 0–1) | Celia · Risk Confidence | 0.88 | The prior run’s confidence in its risk verdict (0–1). |
3. Sample Positive Outcomes query — fields
Same column shape as the In Progress query, with the WHERE clause flipped to capture
successful conversions: WHERE stage IN ('deposited','enrolled'). Celia uses this
to learn what conversion looks like for your institution.
| Field name | Data type | Slate label | Sample value | What this means |
|---|---|---|---|---|
slate_person_ref | Text | — | P-204771 | The institution’s own person reference for a confirmed enrollee. Anonymous identifier; never a name, email, or phone. |
slate_person_guid | Text (UUID) | — | 4a8b1c20-0000-4000-8000-000000000033 | Slate’s internal GUID for the enrollee. Used to dedupe training rows; never PII. |
program | Text | Program | cs_bs | Which program this enrollee committed to. Lets Celia learn program-specific patterns of conversion. |
cohort | Text | Cohort | fall_2025 | Which intake cycle this enrollee belonged to. Last cycle is the most useful training window. |
stage | Text (enum) | Stage | deposited | The terminal positive-outcome stage (typically deposited or enrolled) for this row. Filtered upstream by the query’s WHERE clause. |
stage_entered_at | Date | Stage Entered At | 2025-05-12 | The date this person entered the positive-outcome stage. Pairs with transitioned_at to ground the conversion timing. |
days_in_stage | Number (integer) | Days in Stage | 180 | How long this person has sat in the positive-outcome stage. For training, longer values mean a stable conversion. |
last_touch_days_ago | Number (integer) | Last Touch (days ago) | 3 | Days since the institution last reached out. In the training set this teaches Celia what counselor cadence looks like for converters. |
last_login_days_ago | Number (integer) | Last Login (days ago) | 1 | Days since this person last opened a portal session. Teaches Celia what active engagement looks like for converters. |
is_positive_outcome | Number (0/1) | Is Positive Outcome | 1 | Always 1 on this query — every row is a confirmed positive outcome. Defines the “good” label Celia learns from. |
transitioned_at | Date | Transitioned At | 2025-05-08 | The exact date this person crossed into a positive outcome. Anchors the conversion event for the training window. |
ss_celia_risk | Text (enum) | Celia · Risk | None · Low · Medium · High · Critical | The last risk verdict Celia gave this person before they converted. Used to calibrate the risk model against actual outcomes. |
ss_celia_engagement | Number (integer 0–100) | Celia · Engagement | 82 | The last engagement score Celia gave this person before they converted. Anchors the “what does engaged look like?” baseline. |
ss_celia_readiness | Number (integer 0–100) | Celia · Readiness | 88 | The last readiness score Celia gave this person before they converted. Anchors the “what does ready-to-deposit look like?” baseline. |
ss_celia_yield | Number (decimal 0–1) | Celia · Yield | 0.93 | The last yield score Celia gave this person before they converted. Calibrates the yield model against actual conversions. |
ss_celia_next_action_1 | Text (140) | Celia · Next Action 1 | Maintain orientation contact cadence | The last top-priority action Celia recommended for this enrollee. Useful for narrative training examples. |
ss_celia_next_action_deadline | Date | Celia · Next Action Deadline | 2025-08-01 | The deadline attached to that last action. |
ss_celia_risk_confidence | Number (decimal 0–1) | Celia · Risk Confidence | 0.95 | The last confidence Celia had in the risk verdict (0–1) for this enrollee. |
4. Sample Source Format — fields (30)
The Source Format target accepts the writeback. Declare it on the Person record (most-recent
value, overwritten each run) or on a child Entity record (each run = a new entity,
full history preserved). Same 31 fields either way; the Slate Source Format target alone determines
the behavior. ss_celia_acknowledged_by and ss_celia_ack_outcome are staff-written Slate-side fields — Celia reads
them on the IN query, never writes them — so they are not in the Source Format
target.
| Field name | Data type | Slate label | Sample value | What this means |
|---|---|---|---|---|
slate_person_guid | Text (UUID) | — (match key) | 8f42a91c-0000-4000-8000-000000000000 | Slate’s internal GUID for this person — the match key Celia uses to land the analysis on the right Slate record. Anonymous identifier only; never PII. |
ss_celia_flow_name | Text (80) | Celia · Flow Name | Summer Melt 2026 — CS Cohort | Which Celia Flow produced this row. Lets a counselor tell at a glance which lens is talking when the same person is scored by more than one Flow. |
ss_celia_risk | Text (enum) | Celia · Risk | None · Low · Medium · High · Critical | Celia’s overall risk verdict for this person. Drives the urgency of the next action. |
ss_celia_engagement | Number (integer 0–100) | Celia · Engagement | 78 | Engagement score 0–100. How active this person is right now (logins, opens, replies) relative to their cohort. Higher = more engaged. |
ss_celia_readiness | Number (integer 0–100) | Celia · Readiness | 42 | Readiness score 0–100. How close this person is to being able to make a deposit decision (FAFSA in, transcripts received, blockers cleared). Higher = more ready. |
ss_celia_yield | Number (decimal 0–1) | Celia · Yield | 0.61 | Yield score 0–1. Celia’s estimated likelihood that this person will actually enroll if admitted. Higher = more likely to yield. |
ss_celia_next_action_1 | Text (140) | Celia · Next Action 1 | Send FAFSA reminder today (SMS + email) | The single highest-leverage action a counselor should take next for this person. Short, specific, and verb-first. |
ss_celia_engagement_baseline | Number (integer 0–100) | Celia · Engagement Baseline | 66 | The cohort baseline engagement score (0–100) Celia is comparing this person against — their peer median, derived from the institution’s own positive-outcome history. |
ss_celia_engagement_delta | Number (integer −100–100) | Celia · Engagement Δ | 12 | How far above or below baseline this person’s engagement is, on the same 0–100 scale. Positive = better than peers; negative = behind. |
ss_celia_engagement_conf | Number (decimal 0–1) | Celia · Engagement Conf | 0.84 | How confident Celia is in the engagement score (0–1). Low confidence usually means thin signal — give the score less weight. |
ss_celia_engagement_text | Text (140) | Celia · Engagement Note | ↑ above baseline, cadence fading | One-line counselor-facing note explaining the engagement score. Safe to surface in mailings or counselor dashboards. |
ss_celia_readiness_baseline | Number (integer 0–100) | Celia · Readiness Baseline | 65 | The cohort baseline readiness score (0–100) Celia is comparing this person against, derived from the institution’s own positive-outcome history. |
ss_celia_readiness_delta | Number (integer −100–100) | Celia · Readiness Δ | -23 | How far above or below baseline this person’s readiness is, on the same 0–100 scale. Negative values flag a stall. |
ss_celia_readiness_conf | Number (decimal 0–1) | Celia · Readiness Conf | 0.91 | How confident Celia is in the readiness score (0–1). Pair with the score itself before triaging. |
ss_celia_readiness_text | Text (140) | Celia · Readiness Note | ↓ stalled 23 days, 2 blocking items | One-line counselor-facing note explaining the readiness score. Safe to surface in mailings or counselor dashboards. |
ss_celia_yield_baseline | Number (decimal 0–1) | Celia · Yield Baseline | 0.68 | The cohort baseline yield probability (0–1) Celia is comparing this person against, derived from the institution’s own positive-outcome history. |
ss_celia_yield_delta | Number (decimal −1–1) | Celia · Yield Δ | -0.07 | How far above or below baseline this person’s yield probability is, on the same 0–1 scale. Positive = more likely to yield than peers; negative = less likely. |
ss_celia_yield_conf | Number (decimal 0–1) | Celia · Yield Conf | 0.78 | How confident Celia is in the yield score (0–1). |
ss_celia_yield_text | Text (140) | Celia · Yield Note | high-likely; readiness drag -0.14 | One-line counselor-facing note explaining the yield score. Safe to surface in mailings or counselor dashboards. |
ss_celia_risk_confidence | Number (decimal 0–1) | Celia · Risk Confidence | 0.82 | How confident Celia is in the overall risk verdict (0–1). Pair this with ss_celia_risk before triaging. |
ss_celia_risk_context | Long text (500) | Celia · Risk Context | Historical melt rate for this stage+timeline pattern: 38% if not touched within 7 days. | Free-text explanation of what’s driving the risk verdict — historical pattern, comparable cohort, what tends to go wrong here. Safe to surface to counselors. |
ss_celia_next_action_2 | Text (140) | Celia · Next Action 2 | Personal counselor call within 72h | Second-priority counselor action — what to do after the top-priority one. |
ss_celia_computed_at | Date | Celia · Computed At | 2026-04-23 | The date this scoring run was computed. The “as-of” stamp for everything else in the row. |
ss_celia_data_as_of | Date/Time | Celia · Data As Of | 2026-04-22T23:59:59Z | The cutoff timestamp of the underlying Slate data Celia scored against. Useful when investigating why a score lags reality. |
ss_celia_next_refresh | Date/Time | Celia · Next Refresh | 2026-04-24T02:00:00Z | When the next scoring run is scheduled. Tells a counselor how stale the row is allowed to get. |
ss_celia_model_version | Text (32) | Celia · Model Version | celia-v1.4 | Which Celia model version produced this row. Audit-trail breadcrumb when comparing scores across runs. |
ss_celia_next_action_3 | Text (140) | Celia · Next Action 3 | Nudge recommender on letter | Third-priority counselor action — the next thing after actions 1 and 2. |
ss_celia_next_action_why | Long text (500) | Celia · Next Action Why | FAFSA 14d overdue; cohort median 9d. | Why Celia chose the top action — the one-sentence reasoning a counselor can paste into a personal note. |
ss_celia_next_action_deadline | Date | Celia · Next Action Deadline | 2026-04-26 | The recommended date by which the top action should be taken. Past this date, the action loses leverage. |
ss_celia_signals_explained | Long text (800) | Celia · Signals Explained | Engagement 78 driven by 3 logins/7d. | The longest counselor-facing narrative — what signals are driving the engagement, readiness, and yield scores in plain English. |