Bantr
Rewards creators whose posts on X drive real product usage. I built the frontend and, on the backend, the sentiment scoring and post verification that determine reward eligibility.

- Company
- Intract
- Role
- Frontend Engineer
- Years
- 2025
- Live
- bantr.fun
- Stack
- React, Next.js, TypeScript, X API
Context
Crypto Twitter is loud, and most of the mindshare projects pay for never turns into anyone using the product. Bantr measures which creators actually drive onchain activity, ranks them on public leaderboards and rewards them for it.
What I owned
- Landing page and its scroll-driven hero
- App homepage, stats and leaderboards
- X data ingestion and transformation
- Sentiment analysis and creator scoring
Engineering notes
Remove the noise
The launch hero opens buried under posts from Crypto Twitter. Scrolling clears them away until only the product’s message is left.
From raw posts to a score
Posts arrive raw from the X API. I reshaped them into the data our scoring needed, analysed their sentiment and calculated each creator’s score from it.
Verified before rewarded
Posts are checked for eligibility before they count toward rewards, so payouts follow genuine contributions rather than spam.
Dense data, still readable
Stats, leaderboards and tables built to stay scannable across campaigns, creators and time ranges.












Outcome
Bantr ran live reward campaigns with Web3 projects, ranking creators and paying out against the scores this pipeline produced.
Product design by Mayank