Atlas Platform

Product Classifier

simpleconfidence high — a separate axis from tier; low confidence never auto-decides
Atlas PlatformCataloguehash · sha256:4f1c…9ab2Product Classifierwf-at2d3aaread throughLLAtlas Gatewayhttps://llm.atlas.internal

Where this came from

Every figure below is only as trustworthy as the account it was read through

Application
Catalogue
identified by hash · sha256:4f1c…9ab2
Project
Atlas Platform
Atlas Logistics
Fingerprint
wf-at2d3aa
inferred from logs

Cost

30-day spend

$1,056

LOGGED

On recommended path

$616

MODELLED

Share of attributed spend

7%

Current → recommended (diffed against production)

In production today

gemini-2.5-pro

reasoning high · cap 4,096

Recommended

gemini-2.5-flash-lite

reasoning off · cap 256

  • Δ model: gemini-2.5-pro → gemini-2.5-flash-lite
  • Δ reasoning: high → off

Replay evidence

No measured evidence for this workload.

Replay was not run here, so every quality and cost figure above is modelled or assumed — and labelled as such. This is the common case, not an error.

cannot saysession identifiers do not exist in this log source