Momentor

Momentor is designed to enhance engagement and participation during events by rewarding users for capturing and sharing standout moments. Leveraging blockchain technology, it ensures that content creators receive proper recognition for their contributions, forging a direct link between fans and the originality of their submissions. To add an element of competition and fun, Momentor incorporates AI to assess the quality of submissions, encouraging users to upload top-notch content.

The process is straightforward and engaging: Event organizers list their event on Momentor and set up various bounties aimed at capturing key moments, such as the first goal in a match, a DJ’s set at an NFT club party, or a blockchain hackathon's mascot. Creativity is encouraged, with bounties for unique moments like a team celebrating their project submission. Users, dubbed 'momentors', then submit their photos. Our AI evaluates these submissions for relevance, awarding a score. If a submission meets a set relevance threshold, it triggers the minting of an NFT for the user, marking their entry on the leaderboard for that bounty.

This approach not only fosters a vibrant community but also generates high-quality content for event organizers to make use of assets (i.e. marketing material), training AI models or creating virtual spaces (AR/VR) that recreate the moment, all while still being able to reward and give credit to the original contributors.

Momentor integrates seamlessly with EVM blockchains, requiring minimal setup from both organizers and users. Through a user-friendly interface, participants can upload their content, receive feedback from AI analysis, and earn rewards for their contributions. Organizers can create multiple bounties during an event, selecting their reward (token, amount) of choice, that will be distributed after the event has ended. Likewise, the organizers can view the event media and analyze the user engagement data as well as the leaderboard.

Behind the scenes, each bounty has an associated NFT smart contract that is responsible both for accounting and rewarding each contribution. Once a moment is captured, our backend calls the AI service asking for a quality score of the image. If a certain threshold is satisfied, then the backend uploads the image to Filecoin and mints an NFT to the contributor by calling the smart contract. Since the backend performs everything on the behalf of the user, we don't require any gas transaction to be sent, except when the organizer has to transfer the reward to the contract.

On the AI side, we use an image-to-text model to generate captions out of the image uploaded by the user. Given the caption and the bounty description, we compare the associated vector embeddings to accurately rate the image in terms of relevancy to the bounty before uploading the image to Filecoin using Lighthouse. This means that we are able to generate high quality, labelled training data for the event organizers, which they could then use to understand the fan experience, charactirize their fans to better cater to their interests, or even generate virtual media of the event in the future.