Catalog record · Operations & Efficiency
Stage as of Oct 2, 2025 · historical report, not live reconfirmation.
Internal accounting workflows learned from screen recordings
Primary workflow source. A named organization's workflow supported by a reviewed primary source. Stage is reported separately; this label does not establish measured outcomes or independent validation.
Last researched Oct 26, 2026. Source access and reported stage are recorded below; deployment evidence does not establish ROI.
What the research supports
An internal agent turns accountant screen recordings into reusable instructions for a spreadsheet agent, including revenue reconciliation and intercompany-transfer journal preparation.
- Reported stage
- Pilot
- Technology
- GenAI/LLM
- Business function
- Back Office
- Impact (legacy label)
- Not labeled
- Catalog date · unconfirmed
- Not reported
- Imported
- Oct 29, 2026
- Last researched
- Oct 26, 2026
Measures & evidence
10 · under minutes
Task execution time
Measurement window not disclosed
Scope, baseline & source
- Scope / denominator
- Measurement population / workflow scope not disclosed
- Baseline
- Not disclosed in this account
- Claim basis
- Ramp reports tasks formerly taking 1–2 hours
Self-reported examples; not a formal operational benchmark.
Read dossier source ↗Exact claim-to-source location has not been mapped. This link belongs to the research dossier.
Measures retain their original currency, scope and period. Capacity, adoption and forecasts are not realized cash savings.
Documented vendor relationships
No vendor relationship has been linked to this specific workflow in the curated overlay. This does not mean it was built without vendors.
See the company’s documented relationshipsDeployment evidence & limits
Ramp Labs reports a built internal system, example workflows and experiments; production breadth is not established. Status as of 2025-10-02.
Stage is as reported in the cited evidence. Current operating status has not been independently reconfirmed; the research date is an editorial review date.
No formal Architect benchmark; sampled SpreadsheetBench accuracy was 49.5% on soft-restriction and 32.5% on hard-restriction tasks. Benchmark scores do not measure end-to-end finance correctness.
Recorded dependencies
- Screen recordings, current financial files and human validation.
Source register
- We built an agent to prompt our internal finance agentlabs.ramp.com · opens in a new tab
Ramp Labs · issuer · full-text · Published 2025-10-02
Research lanes: business_finance_digital_assets_gap · Batch global-discovery-2026-10-26. Coverage and remaining gaps