Skip to content
allanninal.dev/build
Collaborate

Part 6 of 7 · Capacity forecaster series ~5 min read

What the capacity forecaster costs

There is no model in this system and the volumes are tiny: a weekly run per production area is twenty runs a month for a business with several. Here is where each cent goes.

Key takeaways

  • About $1 a month at 20 runs. Roughly $1 at 80 runs.
  • One Bedrock read per forecast run is the only line that scales. Everything else is rounding.
  • Nothing is always-on, so a quiet month genuinely costs almost nothing.
  • Storing every forecast at every horizon is the only growing line, and it is kilobytes.
  • The duplicate test runs before the read, so resends are free.
  • The three real risks: a retry loop, storage nobody expires, and a bigger model than the job needs.

The bill at three volumes

These are US East prices at the time of writing, at three volumes that bracket most small businesses. Find the bar closest to your own and read across.

Monthly cost of the capacity forecaster at three volumesA stacked bar chart with three bars, one per volume tier: 4 runs totalling about $1, 20 runs totalling about $1, and 80 runs totalling about $1. Each bar is stacked from bands. The largest and fastest-growing is Bedrock, one read per forecast run, in teal. Then SES for the messages, in pink. Then S3 and DynamoDB storage in green. Then a fixed orange band for Secrets Manager and AWS Budgets, which is eighty-six cents at every volume. Then a grey band for Lambda, SQS and CloudWatch. A note says the read is the only bar that grows with the business and the orange band never moves.$0$0.5$1$1.5$2~$1.244 runs~$1.2720 runs~$1.3780 runsBedrock — one read per forecast runSES — asks, results, receiptsS3 + DynamoDBFixed — Secrets Manager, BudgetsLambda, SQS, CloudWatchThe read is the only bar that grows with the business. The orange fixed band never moves.
Fig 1. The monthly bill at three volumes. The teal band — one model read per forecast run — is the only part that grows; the orange fixed band is the same 86 cents at every volume.

Line by line

LineAt 20 runsHow it scales
Bedrock read$0.00Linear. One call per forecast run, roughly 1,800 in and 200 out tokens.
SES$0.01Linear. About 1 messages per forecast run.
DynamoDB + S3$0.28Storage grows with what you retain, not with throughput.
Lambda + SQS$0.12Linear, and effectively free at this scale.
CloudWatch$0.16Flat, if you set retention. Unbounded if you do not.
Secrets Manager$0.40Flat. One secret, $0.40 a month.
AWS Budgets$0.46Flat. Two actions, so you find out before the bill does.

There is no read line: nothing here calls a model. Messaging is one summary per run, which is why it sits at one per unit.

The three ways this bill surprises you

Every one of these has happened to somebody, and all three are cheap to prevent.

  • Storing only the latest forecast. Not a cost problem: it makes the entire scoring exercise in Part 5 impossible, which is where the improvement comes from.
  • Recomputing on every order change. A forecast is a weekly artefact. Recomputing continuously produces a number that moves for no decidable reason.
  • Building a live utilisation dashboard. The output is a dated warning, not a gauge. A gauge invites watching rather than deciding.

What it costs when nothing happens

This matters more than the headline number for a seasonal business. In a month with nothing to process the bill is the fixed band: Secrets Manager at forty cents, AWS Budgets at forty-six, and a few cents of storage. Call it a dollar. There is no instance to stop and nothing to remember to turn off.

The monthly bill at four volumes plus one failure modeA horizontal row of five boxes. Quiet month, about one dollar. 4 runs, about $1. 20 runs, about $1. 80 runs, about $1. And one bad retry loop, about two hundred dollars. A note says four of these are the design working and the fifth is a missing dead-letter queue.THE BILL, AT A GLANCEQuiet month~$14 runs~$120 runs~$180 runs~$1One bad loop~$200Four of these are the design working. The fifth is a missing dead-letter queue.
Fig 2. The bill at a glance, including the one that is not a volume at all. A retry loop with no dead-letter queue costs more than every legitimate use of the system put together.
  • Management
  • Analytics
  • Front-end & mobile

Set these on day one

  • A dead-letter queue on every SQS queue, with a maximum receive count of three.
  • Thirty-day retention on every CloudWatch log group. There is no default that is safe.
  • An S3 lifecycle rule on the object prefix, tiering at 90 days and expiring at your actual record-keeping horizon.
  • Two AWS Budgets actions — one that emails at half your expected spend, one at double it. The second is how you find out about a loop in an hour instead of a month.
  • Provisioned concurrency: none. Nothing here is latency-sensitive enough to justify paying for a warm function.

Next: the same system drawn for engineers — service names, resource identifiers, IAM scopes, table schemas and the model id.

All posts