---
title: "We built an agent to prompt our internal finance agent"
description: "What happens when you build an AI agent whose job is to figure out how to prompt another AI agent? Lessons from recursive agent architectures."
date: "2025-10-02"
authors: "Ramp Labs"
tag: "Agents"
slug: "agent-to-prompt-finance-agent"
canonical: "https://labs.ramp.com/research/agent-to-prompt-finance-agent/"
---

> [Figure: agent-to-prompt-finance-agent/ascii-hero — ascii-logo]

*What happens when you build an AI agent whose job is to figure out how to prompt another AI agent? Lessons from recursive agent architectures.*

[Ramp Labs](https://labs.ramp.com/) is the home for AI experiments from @tryramp. We share learnings on applying the latest models to real-world problems. We also share insights into how Ramp is leveraging AI internally to increase productivity. Follow us @RampLabs and subscribe to our [newsletter here](https://ramplabs.substack.com/).

[Video: Demo of our internal finance agent](/research/articles/agent-to-prompt-finance-agent/01.mp4)

*Demo of our internal finance agent*

- User uploads screen recording of their workflow
- Gemini 2.5 Pro (1M token context window) processes video to extract actions and context
- Agent identifies required files and temporal dependencies (e.g. current month data)
- Generates structured prompt with step-by-step instructions
- Compiles into a Process that can be reused

Doer agent flow

> [Figure: agent-to-prompt-finance-agent/doer-flow — agent-flow]

- Receives process with instructions and uploaded files
- Claude Sonnet 4.5 uses specialized tools to navigate and manipulate spreadsheets
- Executes each instruction with validation checks
- Returns completed spreadsheet

## Performance results

Architect agent
While we haven’t yet developed a formal benchmark for the architect agent, we can see its impact in terms of workflow creation from non-technical users. Accountants simply upload a screen recording and come back to a ready-to-use process. The time savings are significant when compared with manual prompting for every run.

Doer agent
We tested against [SpreadsheetBench](https://spreadsheetbench.github.io/) on ~50 randomly selected tasks (3 test cases each):

- Soft-restriction tasks: 49.5% accuracy (vs. OpenAI’s 45.5%)
- Hard-restriction tasks: 32.5% accuracy (vs. GPT-4o’s 13.38%)

Tasks taking 1-2 hours now complete in under 10 minutes (a 6-12x speedup). In cases of partial completion, the agent still saves significant time before human intervention is needed.

The barrier to automation collapsed from weeks of engineering work to simply recording your screen.

## Learnings

1. Second-order changes the scaling equation. First-order: Adding workflows scales linearly with human prompt-writing capacity. Second-order: Adding workflows costs near-zero human time once the architect agent exists.
2. Models have complementary strengths. Gemini 2.5 Pro’s 1 million token context window with stellar multimodal support allows us to capture key details from input videos to generate a process, while Anthropic’s Claude Sonnet 4.5 shines when it’s time get work done by being swift, precise, and economic.
3. Building production agents is still hard. Despite powerful models, we spent weeks iterating on system prompts, tool designs, and error handling. There’s no shortcut for testing, debugging, and refinement. The engineering matters as much as the models.

Our combined agent doesn’t just do work, it designs how work gets done. This unlocks a new scaling paradigm: instead of engineers building automations one by one, anyone who can record their screen can generate reusable workflows. The architect-doer pattern transformed automation from an engineering bottleneck into a self-service capability. The next frontier isn’t faster agents, it’s agents that teach themselves what to do.

Want to get early access to future AI experiments? Subscribe [here](https://ramplabs.substack.com/) and follow us on @RampLabs. We’re also [hiring across roles](https://jobs.ashbyhq.com/ramp?utm_source=RampLabs) at Ramp.

## Citation

## Citation

See article metadata for citation details.