Inspiration
As a group of students who admittedly don't go outside very often, we thought about products and ideas that would help us become more familiar with both outdoor activity and Arc'teryx products. ๐ณ
Our project is perfect for anyone looking to get outside, simplifying the amount of information and research needed to set off on their adventures while encouraging them to make purchases that suit their use-cases. We aim to build a product fit-to-purpose and accessible to people of all familiarity to trails and outdoor activities.
Oniva is derived from the french translation of "Let's Go!" to encourage everyone to take a chance and head outside and enjoy local trails.
What it does
We aim to solve both trail and product discovery in one simple product discovery tool. Our Oniva activity finder allows users to get recommended a trail according to either activity preferences or the gear they currently own. Oniva pairs the recommended activity with matching Arc'teryx products, allowing users to seamlessly make purchases. This provides a personalized recommendation process, while directly encouraging and providing options for outdoor exploration. โฐ๏ธ
In consideration of Omnichannel integration, we have aligned Oniva to correspond to existing Arc'teryx collection naming conventions.
How we built it
Our Tech Stack: We chose the MERN tech stack primarily because of prior experience with the stack. On the Frontend side Tailwind CSS was used for styling and Vite for setting up local development servers and tooling.
Additional tools include the Google Cloud Vision API, which was used to analyze clothing types and whether clothing choices are appropriate for given outdoor activities. Confidence scores for specific labels were factored into an overall confidence value, which would inform the score seen on the website.
Challenges we ran into
- We struggled to find a database that included local trails, in result we had to create our own database manually which took a lot of effort and time.
- Our original geolocation mechanism was made using Google Maps Nearby Places API, however we ran into challenges of getting user geolocation data and running the comparison based on only the address JSON field.
- We didn't have access to Arc'teryx product information/database, so we couldn't use Google's Product Search API. Instead, we ended up just relying on their default labelling system from Google Vision AI and scaled, not off specific activity, but on general "sports" classification confidence.
- We struggled to figure out how to render our results page multiple times according to the data being processed.
Accomplishments that we're proud of
- First time using Google Cloud Vision and Google Maps API! โ๏ธ
- Mentor got stuck with us on our backend (hey, it wasn't just us struggling...) ๐ฃ
- Alignment with Arc'teryx's design principles/aesthetic. ๐ฉโ๐จ
What we learned
It is extremely difficult to create full-stack projects with multiple API integrations. You can't always find the databases you need. Google Cloud APIs are good and extremely easy to understand via their documentation. Hackathons are stressful but good teammates make it better ! ๐ฅ
What's next for Oniva
- Implementation of a streak/rewards system (gamification) ๐ฎ
- Connections with Arcโteryx customer account
- Support larger range of outdoor activities
Log in or sign up for Devpost to join the conversation.