Shape a small product improvement in v0, hand the refined specification to Devin, and use the working preview to finish the feature quickly.
Source episode · 04:00
How I AI: Gumroad CEO Sahil Lavingia’s Workflow for Building Features 40x Fasterwith Claire Vo
04:00 to 29:15: Sahil Lavingia spots a weak date picker, refines the interaction in v0, sends the resulting specification to Devin, and inspects the implementation in a preview.
A pull request and working preview for a focused UI improvement whose interaction was explored before implementation.
AI-first code editor
Vercel AI UI generator
AI software engineer by Cognition Labs
Step by step
Follow the sequence once, then adapt the prompts, checks, and handoffs to your own setup.
5 steps
Choose a concrete irritation that is small enough to ship but meaningful to the user. Sahil starts with a native date input that cannot understand phrases such as "next Monday."
On [route or screen], replace the current [component] because users struggle with [specific friction]. The improved version must support [core behavior], [important edge case], and [mobile or accessibility requirement]. Use our existing [design system or component library].
Build the interaction in v0 and keep refining it beyond the first pass. Try realistic inputs, add useful controls, and settle the details that would otherwise become unanswered implementation questions.
Design a natural-language date picker for an HR onboarding form. Let people type phrases such as "next Monday" or "in one hour," show the parsed date before saving, include clear and delete behavior, and use shadcn/ui. Keep the interaction compact and accessible.
Give Devin the refined requirements, the target repository and route, and the implementation constraints. Share the v0 result as a reference, but describe the behavior in text so the task is not dependent on an image.
In the [repository] repo, update [route or component]. Replace the browser-native date input with a natural-language date picker using our existing shadcn/ui components. Requirements: - Accept dates such as "tomorrow," "next Monday," and "three Sundays from now" - Show the parsed date before submission - Include clear and keyboard behavior - Preserve the existing form contract and styling Run the app, test the interaction in the browser, and open a pull request with the result.
Let the agent set up the project, find the relevant files, and implement the feature. Use its progress and preview to catch setup errors or misunderstood requirements while the work is still in motion.
Before finishing, run the project and exercise the component with: "tomorrow," "next Monday," an invalid phrase, clearing the field, and keyboard-only input. Fix any failures and summarize the files and tests changed.
Open the preview and use the feature as a customer would. Inspect the code where product naming, architecture, or edge-case behavior matters, then send any fixes back to the agent or finish them in Cursor.
Update the pull request for these preview findings: [list findings]. Add or update the smallest useful tests, confirm the original form still submits the same value shape, and explain any behavior you intentionally left unchanged.
Turn an idea into a PRD, user stories, and a plan.
After the steps
How to recover when the loop fails and where human judgment helps.
Recover
Keep building
Log your workouts effortlessly without ever opening a spreadsheet. This workflow uses the Claude mobile app to parse a simple voice note and automatically update your workout tracking spreadsheet in the background.
Create a custom email management interface within Claude to read, triage, and draft replies without opening the Gmail tab. This workflow helps you reclaim focus and compound your learnings by training your AI on every interaction.
Build a fully automated system in Claude to generate personalized, interactive HTML proposals and client onboarding materials, saving hours on administrative work and creating a high-touch experience.
Join 100,000+ product managers who use ChatPRD to write better docs, align teams faster, and build products users love.