MCP use cases

Business OperationsLocal Business

Use agentic workflows to help small businesses move from sales data to action: analyzing performance, updating a store, creating payment links, and completing operational tasks inside the Square ecosystem.

Outcome: Shows how AI agents can support small-business operations by connecting analysis to execution. Instead of only summarizing sales data, the workflow can help identify what needs to change, update business systems, and create the next customer-facing action.

Prototyping & Design

Built ProtoDash, an AI-powered internal prototyping studio that brings Stripe’s design system into AI-generated prototypes, helping designers and PMs create high-fidelity clickable prototypes instead of generic AI output.

Outcome: Turns AI prototyping from generic “blurple slop” into on-brand, shareable Stripe prototypes; supports live demos instead of static memos, helps PMs prototype, and streamlines design feedback and engineering handoff.

Prototyping & Design

Uses structured JSON data and a clear prompt framework to make AI prototypes and Midjourney images more realistic, flexible, and production-useful.

Outcome: Moves AI prototyping beyond vague “make it look nice” prompts. By separating data, UX, and visual style, Ravi gets prototypes that feel more like real products and images that look more curated than generic AI output.

Coding & EngineeringBusiness Operations

Helped build Stripe Minions, an internal AI coding-agent workflow where employees can trigger agents from Slack, have them make code changes, and review the resulting pull requests before anything ships.

Outcome: Turns common engineering requests into a lightweight agent workflow inside Slack. Instead of opening an editor for every small change, teams can dispatch an agent, review its PR, and keep humans responsible for approval.