GearSmart

Project Description

GearSmart is a data-powered gear advisor that builds a trip-specific packing list for people heading outdoors for the first time. The idea came from a simple gap: the moment before a first trip when you don't know what you don't know. I'd lived it on my own first trips and watched it constantly working the floor at REI and meeting beginners through the Ladies Climbing Coalition. There's no shortage of gear content online. However, there's a shortage of gear content relevant to a specific beginner on a specific trip. GearSmart digitally recreates the experienced friend who already knows what's needed. A user enters their destination, dates, activity, experience level, and budget; the product generates a categorized checklist where every item carries a plain-language "why," tied to live weather for that location. It's a fully deployed, security-hardened product, validated with real beginners, an industry panel, and a public exhibition. This was a solo capstone where I led the research, product design, schema design, and AI-assisted engineering end to end.

Problem

Every year thousands of people head outdoors for the first time underprepared — not for lack of trying, but because the available information was built for someone who already knows. Research and store-floor experience surfaced three root causes. Condition gaps: generic lists don't account for where you're actually going and beginners can't anticipate what they've never encountered. Cost intimidation: walking into a gear store as a beginner is overwhelming, with $600 tents and $200 sleeping bags, so first-timers either overbuy or check out entirely. Information overload: the internet is saturated with gear advice, but almost none of it is filtered to one specific person on one specific trip. The deeper problem I uncovered during testing was subtler still: beginners consistently over-rate their own readiness on exactly the trip where that gap matters most. The product had to close all three gaps without adding to the noise.

Insights

I anchored the research in the Ladies Climbing Coalition, real beginner outdoor women, and other outdoorsmen of various experience levels through three "tiny experiments" and five distinct methods. (1) Field research: observing beginners at LCC meetups and on the REI sales floor. (2) A "sommelier" gear roleplay: I laid out a full gear spread and presented pieces one at a time the way a seasoned employee would, so participants externalized what they'd never think to ask. (3) An affinity board: a "What do you wish you knew?" wall where participants posted sticky notes that clustered bottom-up into condition surprises, gear paralysis, and specificity gaps. (4) Personas built from those patterns, and (5) iterative prototyping and testing across four versions (from an early Magic Patterns sketch to the live build) with remote testers and exhibition observation. The notes spoke for themselves: "It's REALLY cold at night in the mountains/desert," "Climbing/camping gear can get VERY expensive," and "Literally everything — no idea what to bring." One beginner, asked what would have helped, wrote: "Having a very knowledgeable and well-equipped and generous friend," which independently named the exact metaphor GearSmart is built on. The biggest insight was calibration: the real problem wasn't gear, it was beginners over-rating their own skill on the trip where it matters most.

Solutions

GearSmart turns a short trip questionnaire, asking for destination, dates, activity, experience level, and budget, into a curated and categorized gear checklist. Each item carries a plain-language reason it's on the list, and "Essential" badges plus a completion progress bar replace anxiety-inducing flat lists with a clear, finishable structure. A live weather card surfaces the counterintuitive conditions beginners miss, like freezing desert nights. Rather than a static database, the product uses AI to generate genuinely trip-specific lists, because no fixed catalog could cover every destination-by-month-by-activity-by-experience-by-budget combination. I used a retailer-agnostic price router that finds the most trustworthy low price across six retailers. An AR Fit Guide coaches users through fitting a backpack in five steps. Behind the scenes I re-architected the app to move API keys server-side, harden it against errors, and cap spending, so that it's a real product and not just a demo.