One row per person, from the users table. Sessions counts
the conversations their questions were grouped into; the count links through
to them. (unattributed) collects questions with no
user_id: fail-soft writes, and history migration 013 could not
match to a person. Those questions never get a session either, so that row
always reads zero — they are listed from the Sessions tab instead.
| User | Identity | Sessions | Questions | Cost | Cost / question | Unpriced | Success | Feedback | Last seen |
|---|---|---|---|---|---|---|---|---|---|
|
yordan@dext.com
Yordan Todorov Ivanov
|
14 | 34 | $265.0288 | $7.7950 | 0 | 94.1% | — | 2026-09-18 10:51 | |
|
gerty.bester@dext.com
Gerty Bester
|
3 | 5 | $172.4565 | $34.4913 | 0 | 100.0% | — | 2026-09-29 10:45 | |
|
chris.scott@dext.com
Chris Scott
|
5 | 76 | $112.0027 | $1.4737 | 0 | 98.7% | — | 2026-09-22 07:23 | |
|
ikraan.mohamed@dext.com
Ikraan Mohamed
|
1 | 1 | $58.3348 | $58.3348 | 0 | 100.0% | — | 2026-09-22 14:51 | |
|
suzannah.weinfass@dext.com
Suzannah Weinfass
|
3 | 4 | $22.4246 | $5.6061 | 0 | 100.0% | — | 2026-09-28 14:05 | |
| silviya.chomakova@dext.com | 1 | 16 | $14.2878 | $0.8930 | 0 | 100.0% | Positive | 2026-07-31 11:31 | |
| detelina.vassileva@dext.com | 2 | 8 | $14.0686 | $1.7586 | 0 | 87.5% | — | 2026-08-06 10:20 | |
|
sukh.kalsi@dext.com
Sukh Kalsi
|
2 | 3 | $2.3310 | $0.7770 | 0 | 100.0% | — | 2026-08-13 11:46 | |
|
lindsey.kistler@dext.com
Lindsey Kistler
|
1 | 2 | $1.6985 | $0.8492 | 0 | 100.0% | — | 2026-08-03 10:45 | |
|
kathryn.buckley@dext.com
Kathryn Buckley
|
1 | 2 | $1.1993 | $0.5997 | 0 | 100.0% | — | 2026-08-17 09:54 |
Cost is the LLM completion spend LiteLLM priced for each call, summed per
request.
Chandler is running on the Codex CLI, which draws ChatGPT plan quota
rather than per-token API billing, so this figure is not money paid —
it is what the same traffic would have cost on the API, priced from the
token counts Codex reports. Read it as the size of the bill avoided.
Two known limits: embedding spend is not recorded, so
retrieval and matching cost is missing, and because the figure is one sum
per request it cannot be split by model within a request — a
request's classifier and agent calls can use different models while
llm_model holds only one name.
Only authenticated calls are logged, USAGE_LOG_ENABLED can
switch logging off, and log writes are fail-soft — this is not a complete
record of traffic. Questions are grouped into sessions: a session is
exact when the caller echoed its id back to us and otherwise inferred from
a 30-minute gap in that user’s activity, so a grouping is only as
good as the source shown on the session itself. A call with no
attributable user gets no session at all; those questions are listed
separately rather than dropped.