Canonical: https://botbento.com/usecases/check-sales-transcript-budget-figures-for-arithmetic/
Format: Markdown representation of the public HTML page.

# Check sales transcript budget figures for arithmetic conflicts

[Operations](https://botbento.com/usecases/ops/) · Intermediate

Extract numeric claims and flag mismatches before handoff

A bot reviews sales call transcripts and pulls out budget, price, quantity and total figures. It checks whether the stated numbers satisfy simple arithmetic and business rules. It records confidence and cites the transcript lines behind each value, flags contradictions for a person, and prepares a clean review note for downstream use.

## Quick answers

**What does it do?** Extract numeric claims and flag mismatches before handoff. Result: Review note with flags.

**What does it need?** Tools: Docs, Spreadsheet, Chat. Intermediate: some setup.

**How do I run it?** When something happens: New transcript arrives. Copy the prompt below into a bot in BotBento.

## How it works

1. When something happens: New transcript arrives
2. Then: Extract and verify figures
3. Result: Review note with flags

## Prompt to start with

```text
When a new transcript is available, read it and extract every stated budget, price, quantity, and total value. For each value, record the exact line or phrase that supports it and a confidence level. Check whether total equals price times quantity and whether other stated rules hold. If a value is missing or contradictory, mark it for review instead of guessing. Prepare a short review note listing accepted values, flagged values, and questions for a person. Do not send messages, make payments, delete records, or update any downstream system until the user approves.
```

## Tools it uses

Docs, Spreadsheet, Chat

## Works with

BotBento

## Tags

verification, sales, data-quality, transcripts, approval

## Source

[Aura-State: Formally Verified LLM State Machine Compiler](https://news.ycombinator.com/item?id=47209315) — Hacker News · checked 2026-11-13

Summarized in our own words. Any results described there are the authors’ own, not measured by BotBento.

Verified 2026-11-13: passed every automated gate and a second-model review (kimi-code/k3.1-pro). Automated, not human review. [How verification works](https://botbento.com/usecases/verification/)

By BotBento Editorial · Added 2026-11-13 · Updated 2026-11-13 · [Suggest a correction](https://botbento.com/contact/)

## Related use cases

- [Audit a developer-tool directory for stale or unverifiable claims](https://botbento.com/usecases/audit-a-developer-tool-directory-for-stale/) — Cross-check each listing against its own docs and flag the drift
- [Reconcile linked business tables and flag broken records](https://botbento.com/usecases/reconcile-linked-business-tables-and-flag-broken/) — Cross-check related tables for missing keys and mismatched totals
- [Cross-check claimed archaeological finds against public lidar records](https://botbento.com/usecases/cross-check-claimed-archaeological-finds-against-public/) — Triage new site claims before they reach a research queue
