Flow run detail

What it is

A single Flow run's complete picture. Use it to debug a failed run, audit what was written back, or confirm a successful run actually scored what you expected.

What you can do here

• Read the summary tiles — duration, students analyzed, Writeback rows landed, average Engagement / Readiness / Yield scores.
• See the per-student outcomes — anonymous IDs with their resulting scores across all three axes.
• Read errors — when status is failed or partial.
• Check baseline source — a badge showing which Slate surface(s) provided the positive-cohort reference data Celia was anchored against for this run. Four states:
  • Both surfaces — In Progress positives and the Positive Outcomes query both contributed converters.
  • In query only — converters came from is_positive_outcome = 1 rows in the In Progress query; the Positive Outcomes URL was absent or empty.
  • Positives query only — the dedicated Positive Outcomes query contributed; no converters were flagged in the In Progress query.
  • Empty — no converters found on either surface; Celia scored against fleet defaults with no institution-specific learning.
• Inspect raw metadata — for support diagnostics.

Common workflows

Diagnose a failed run

1. Open the failed run from /schedules or the dashboard.
2. Read the Errors section — most errors are Slate-side (auth expired, query offline).
3. Fix in Slate. Use Run now on the parent Flow to retry.

Confirm the Writeback landed

1. Open the run.
2. Pick a student row.
3. Open them in Slate and verify the ss_celia_* fields show the expected values.

Common gotchas

Throughput is zero but status is "completed"

The In Progress query returned no rows that passed Celia's pre-flight no-PII checks. Verify the query's filters, then re-run.

"Partial" status

Some students scored, some hit a per-row error. The Errors section lists the failed IDs. Most partial runs are caused by missing fields — populate them in Slate and retry.

Run fails with missing_required_fields

One or more of the nine required fields were absent from your Slate query at runtime. Open the parent Flow's settings, confirm all nine fields are in the query output, click Refresh fields, and retry the run. Required fields: student_type, program_interest, entry_term, campus_visit_count, last_activity_date, application_status, home_state, docs_missing, residency_country.