How many accounts have improving sentiment?
Per-step detail — each LLM call, each Omni request, each rejected draft —
is traced to stdout only and is not stored, so it cannot be shown here.
Query attempts counts run_query executions, not model turns.
WITH ranked_responses AS (
SELECT
omni_dbt__eds_all_accounts."ACCOUNT_CRN" AS account_crn,
omni_dbt__prepare_intercom_nps_scores."USER_CRN" AS user_crn,
omni_dbt__prepare_intercom_nps_scores."RESPONSE_SCORE" AS response_score,
ROW_NUMBER() OVER (
PARTITION BY
omni_dbt__eds_all_accounts."ACCOUNT_CRN",
omni_dbt__prepare_intercom_nps_scores."USER_CRN"
ORDER BY omni_dbt__prepare_intercom_nps_scores."OCCURRED_AT" DESC
) AS response_rank
FROM PROD."PREPARE_INTERCOM_NPS_SCORES" AS omni_dbt__prepare_intercom_nps_scores
INNER JOIN PROD."EDS_ALL_ACCOUNTS" AS omni_dbt__eds_all_accounts
ON omni_dbt__eds_all_accounts."DERIVED_ACCOUNT_ID" = omni_dbt__prepare_intercom_nps_scores."ACCOUNT_CRN"
and CAST(omni_dbt__prepare_intercom_nps_scores."OCCURRED_AT" AS DATE) between omni_dbt__eds_all_accounts."VALID_FROM" and omni_dbt__eds_all_accounts."VALID_TO"
WHERE omni_dbt__prepare_intercom_nps_scores."RESPONSE_SCORE" IS NOT NULL
AND COALESCE(omni_dbt__eds_all_accounts."IS_DEXT_DEMO", FALSE) = FALSE
),
user_changes AS (
SELECT
account_crn,
user_crn,
MAX(CASE WHEN response_rank = 1 THEN response_score END) AS latest_score,
MAX(CASE WHEN response_rank = 2 THEN response_score END) AS previous_score
FROM ranked_responses
WHERE response_rank <= 2
GROUP BY account_crn, user_crn
HAVING COUNT(*) = 2
)
SELECT
COUNT(DISTINCT CASE
WHEN latest_score > previous_score THEN account_crn
END) AS "Accounts With Improving Sentiment"
FROM user_changes
LIMIT 1
No feedback linked to this request.
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.