Meet Ramp Research: Our Agentic Data Analyst
F
Faiz Hilaly, Cesar Duran, Jay Sobel 7
Medium Write Soon
This article strongly resonates and has clear angles for your perspective
Quick Take
This hits Brian's sweet spot of AI agents solving real engineering problems with concrete metrics (10-20x question increase, 1,800+ answered). The technical implementation details around context management and agentic tooling directly apply to his fintech work, and the "automating workflows" angle could spark ideas for his side projects.
Relevant Domains
AI/agents/future of software work ✅ (Primary) Engineering craft/architecture/productivity ✅ Side projects/automation/earning from skills ✅
Blog Angles
1
"Why Every Fintech Needs an Internal Data Agent (And How to Build One)"
Thesis
Your Hook
2
"The Real Cost of 'Just Ask in #help-data'"
Thesis
Your Hook
3
"Building AI Agents That Actually Ship: Lessons from Production"
Thesis
Your Hook
Key Quotes
10-20x increase in the number of questions people ask. Most of that growth comes from questions that previously died in drafts or never left someone's head.
Like counting cards, a one- or two-point lift in decision quality doesn't show up in the margin, but spread across thousands of pricing tweaks, GTM filters, and feature rollouts, it becomes material.
Rather than evaluating each question, we shifted to evaluating our context layer.
We built a Python mini-framework in our dbt project. It asserts not only on the final answer but also the intermediate steps
Tags
#ai-agents
#fintech-engineering
#data-automation
#slack-integrations
#agent-architecture
#productivity-multipliers
#internal-tools