One maintained catalogue
Three category finders draw from 373 keyboards, 86 mice and 61 headsets, with automated checks guarding catalogue health and recommendation coverage.
Guided commerce / Recommendation engine
Better recommendations start with the person, not the product list.
PickMySetup replaces generic best-of lists with guided keyboard, mouse and headset finders. Each one weighs practical fit, compatibility, features and budget, then explains the strongest match and the trade-offs behind it.
Repository verified / 24 August 2026



Current proof
03 outcomes / repository verified
Three category finders draw from 373 keyboards, 86 mice and 61 headsets, with automated checks guarding catalogue health and recommendation coverage.
Each finder returns one strongest match and four deliberately varied alternatives, with the reasoning and compromises left visible.
Keyboard, mouse and headset results can be compared, saved together and shared through a privacy-safe snapshot without forcing an account.
The problem
A universal best keyboard, mouse or headset does not exist. People have different hands, desks, devices, environments, budgets and non-negotiable compatibility requirements.
Most buying guides flatten those differences into a ranked list. That makes the recommendation easy to publish, but difficult for a person to trust or apply to their own setup.
The response
PickMySetup turns those differences into a guided decision. Three focused finders ask only the questions that change the outcome, score a maintained catalogue and explain why the best result won.
The platform keeps viable alternatives visible, supports side-by-side comparison and lets a visitor save or share a complete setup without requiring an account.
Product decisions
04 principles / visible in the product
Hard compatibility, ergonomic needs and firm budgets carry more weight than cosmetic or flexible preferences.
Each result explains the match and its compromises so the recommendation can be challenged instead of accepted blindly.
The top five is designed to surface meaningfully different options rather than five near-identical substitutes.
Retail availability can help someone act on a result, but it never changes the score or shortlist position.
What shipped
System view
A React and TypeScript application shares scoring rules across three finders. Netlify hosts the Vite build and server functions, Identity provides optional accounts, Blobs stores saved and shared setups, and isolated previews plus automated checks protect production releases.
The goal is not to name a universal winner. It is to make a specific decision easier to understand.Open PickMySetup
Next file / 02
CardScope