Support agent on live systems architecture
A support agent that reads the orders database and writes back through an approval gate. Agent steps and the systems they touch on one graph.
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.
| Node | Type | What it is | Projected |
|---|---|---|---|
| Orders database | Database | db.t4g.large · multi-AZ · 100 GB | $215/mo |
| CRM (external) | External | existing subscription | no direct cost |
| Evidence archive | Storage | 100 GB standard | $4.00/mo |
| Inbound ticket | Trigger | Helpdesk webhook | — |
| Classify and route | Agent step | Intent, urgency, product area | billed per token |
| Draft reply | Agent step | Grounded in order history and policy | billed per token |
| orders_lookup | Tool | Reads the orders database | — |
| crm_write | Tool | Writes the resolution back to the CRM | — |
| Approval gate | Guardrail | Refunds over $50 need a human | enforced at runtime |
| Budget cap | Guardrail | Halts runs over $120/mo | enforced at runtime |
| ArcFoundry runtime | Runtime | — | — |
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.
Similar templates.
Agent with enforced write control
The pattern for letting an agent touch production: every write passes an approval gate, every action lands in a hash-chained evidence log.
Retrieval service (RAG)
An answer endpoint over your own documents: vector store, retrieval step, generation step, cached responses.
Document processing pipeline
Files land in storage, an agent extracts structured fields, a person approves anything low-confidence, and the result is written to the database.
Open Support agent on live systems on the canvas.
It loads as an editable graph. Connect an account or instrument an agent and the projected figures above become measured ones.