What the timesheet validator costs
Nothing here is always-on, so the bill is almost entirely per-sheet. The one line worth watching is Textract, and it only bills on the photo lane — a business where everybody uses the form pays nothing for it at all. Here is where each cent goes.
Key takeaways
- About $3 a month at 200 sheets. Roughly $8 at 900 sheets.
- One Bedrock read per sheet is the only line that scales. Everything else is rounding.
- Nothing is always-on, so a quiet month genuinely costs almost nothing.
- Textract only bills on photographed sheets. A team that uses the form pays nothing for it.
- 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.
Line by line
| Line | At 200 sheets | How it scales |
|---|---|---|
| Bedrock read | $0.82 | Linear. One call per sheet, roughly 1,800 in and 200 out tokens. |
| Textract | $0.42 | Linear at $0.0021 per sheet. |
| SES | $0.13 | Linear. About 1.4 messages per sheet. |
| DynamoDB + S3 | $0.31 | Storage grows with what you retain, not with throughput. |
| Lambda + SQS | $0.12 | Linear, and effectively free at this scale. |
| CloudWatch | $0.16 | Flat, if you set retention. Unbounded if you do not. |
| Secrets Manager | $0.40 | Flat. One secret, $0.40 a month. |
| AWS Budgets | $0.46 | Flat. Two actions, so you find out before the bill does. |
The Textract line assumes roughly a third of sheets arrive as photographs, which is typical for a mixed office-and-site workforce and pessimistic for an office-only one. Businesses where everybody uses the form can delete that band entirely.
The three ways this bill surprises you
Every one of these has happened to somebody, and all three are cheap to prevent.
- A retry loop on the read. A photo that Textract cannot parse makes the function throw, the retry throws, and the queue redelivers. Without a dead-letter queue that is one bad photo costing more than a hundred good sheets. A maximum receive count of three fixes it.
- Re-reading on every resubmission. If the draft-replacement logic re-runs Textract and the model on an unchanged image, a person correcting one cell four times pays for four full reads. Key the read cache on the image digest, not the sheet id.
- Log retention left at never. CloudWatch keeps log groups forever by default, and on a system this small the logs will eventually cost more than the compute. Thirty days is a one-line change.
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.
- 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