Bulk classification architecture

High-volume, low-cost classification over a queue, with the cheap model doing the work and a cap on the bill.

AWSclassificationhaikuhigh volumequeue
Input bucket100 GB standardClassification queue5M requestsBatch enqueuedQueue messageClassifySingle-label,schema-enforcedBudget capguardrailArcFoundry runtimeruntimeLabels storeon-demand · ~5Mwrites, 20M reads
7 nodes — 3 infrastructure and 4 agent, on one graph. Drawn by the same layout the product uses.
What is in it

Every resource, and what it costs.

Projections from August 2026 list prices for always-on resources. Connect an account and these become the figures your provider actually bills.

Resources in the Bulk classification template with projected monthly cost.
NodeTypeWhat it isProjected
Input bucketStorage100 GB standard$3.00/mo
Classification queueMessaging5M requests$2.00/mo
Labels storeDatabaseon-demand · ~5M writes, 20M readsfrom $15/mo
Batch enqueuedTriggerQueue message
ClassifyAgent stepSingle-label, schema-enforcedbilled per token
Budget capGuardrailHalts runs over $600/moenforced at runtime
ArcFoundry runtimeRuntime

Agent steps are priced from provider-reported token usage once the agent runs, not estimated. Guardrails cost nothing and are the reason a runaway agent cannot. One row shows from because those services bill by usage, so the total is a scenario at the stated volumes, not a quote.

Open Bulk classification on the canvas.

It loads as an editable graph. Connect an account or instrument an agent and the projected figures above become measured ones.