Chandler · usage
← Back to sessions

Session

User
gerty.bester@dext.com · Gerty Bester
Started
2026-09-27 14:47
Last activity
2026-09-27 14:58
Questions
4
Grouping
client — exact — the caller echoed the session id back to us
session_id
f00d6c66-f511-4150-9295-0cce7ff2a5f3

Transcript

  1. 2026-09-27 14:51 ask success $67.8887 205.6 s full detail →
    Total number of AI Assist suggestions created and applied to date (all time, global), and the number of partner practices and client businesses that have had at least one AI Assist suggestion.
    Data answer — see full detail.
    1 row Product Usage - Users
  2. 2026-09-27 14:51 ask success $13.3717 74.6 s full detail →
    How many distinct parent practices currently have at least one client account on Dext Solo? Latest snapshot please.
    Data answer — see full detail.
    1 row Product Usage - Accounts
  3. 2026-09-27 14:58 ask success $38.8849 139.3 s full detail →
    How many AI Assist suggestions were applied in total up to today?
    Data answer — see full detail.
    1 row Product Usage - Users
  4. 2026-09-27 15:01 ask success $29.1916 149.2 s full detail →
    For the same basis, how many distinct accounts had at least one AI Assist suggestion applied, split by partner practices versus client or direct businesses, and how many suggestions were generated (created) in total?
    Data answer — see full detail.
    1 row Product Usage - Users

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.