Snapshooter and Scoring

Snapshooter is the core engine of Miraya. This is designed to run at a certain frequency each day (once/day for now). This fetches the activity of the current campaign-specific hashtag and runs an AI model over it to score all users and rank them accordingly.

Snapshooter is built over ai16z's Eliza framework to effectively fetch data from X, compute with AI, and deliver posts through Miraya's official X handle.

Snapshooter is deployed on Spheron, leveraging decentralized computing and enabling payments with Miraya’s native token for greater feasibility.

It uses:

• X API

• Eliza Framework

• Spheron servers

Miraya's AI model has been assigned specific weights for different insights:

Metric	Weight/Point Calculation
Reach	1 point per 100 views
Likes	1 point per like
Reposts	3 points per repost
Comments	2 points per comment

However, in addition to the above, it also reads the text of the post and assigns points:

• Quality of the text: A fractional value between 1 and 2, based on its insightfulness

The Total Score is calculated as:

$$Total Score=Quality×Point Factor$$

Point Factor:

$$Point Factor=Reach Points+Like Points+Repost Points+Comment Points$$

This ensures that both views and quality are given equal weight. If a bot spams a high-quality post, the engagement will lower its overall score. Conversely, if a post has fewer impressions but offers high value, it will earn more points.

Each user is limited to 3 posts per day to prevent spam.