Aglet

Investigate Why Data Counts Do Not Match

Investigation should turn two totals into comparable sets and operations. Re-run both counting paths over the same snapshot or bounded window, compare identifiers before aggregation, and inspect filters, joins, distinctness, and timing. Treat every difference as an observation to explain, not a presumed defect.

Build a useful investigation brief

  1. Freeze comparable inputs

    Choose a bounded source snapshot or exact event window and record its timestamp, timezone, processing revision, and tenant or entity scope. Run both paths against that same input. If a shared snapshot is impossible, document the timing difference and its effect.

  2. Compare set operations

    Materialize included identifiers for each path, then compute missing, extra, duplicated, and merged groups. Trace each set difference to a filter, join, status interpretation, timestamp predicate, or distinct operation. Preserve query text and intermediate counts so the explanation is reproducible.

  3. Test the smallest change

    Alter one suspected boundary at a time, such as a date endpoint, status predicate, join key, or grouping field. Compare the resulting set and total with the original. If several changes are required, keep their effects separate and state which evidence remains inconclusive.

What to carry forward

Produce a reconciliation brief with frozen inputs, intermediate sets, changed predicates or joins, and the smallest explanation supported by evidence. Recommend a scoped correction or definition decision. Do not declare agreement from matching totals when the included identifiers still differ.

Keep the decision with the work.

Use a Work Item in Aglet to record the problem, the evidence you have, and the next decision. Add an owner and priority, then keep updates in the discussion so the next person can follow the reasoning.

Create an account See the product workflow