CryptoLensAI

Trust scores for any cryptocurrency-explainable AI scoring across on-chain transparency ,team credibility, community authenticity and legal compliance in seconds.

Comment

I spent years in banking and finance — Axis Bank, ICICI Prudential — watching retail investors make decisions based on hype rather than fundamentals. Crypto amplifies this problem massively. The barrier to launching a token is near zero, but the tools to evaluate one remain inaccessible to ordinary people. Whitepapers are long, on-chain data is technical, and social sentiment is noisy. Scams and rug pulls cost retail investors billions every year — not because the information doesn't exist, but because it's scattered and unreadable. I wanted to build the kind of due diligence tool institutional investors take for granted, but for anyone.

What it does

CryptoLensAI analyzes any cryptocurrency and returns a single explainable Trust Score (0–100) in seconds, broken into four transparent dimensions:

  • On-Chain Transparency (0–25) — blockchain explorer availability, genesis date, public notices
  • Team Credibility (0–25) — GitHub activity, commit history, developer engagement
  • Community Authenticity (0–25) — Reddit, Telegram, and social presence, with an intelligent fallback for coins whose communities live outside tracked platforms
  • Legal & Registration (0–25) — exchange listings, market cap rank, regulatory signals

Every dimension comes with a 2–3 sentence plain-English explanation of why that score was given — no black box, just transparent reasoning grounded in real data. The app also surfaces official links (website, whitepaper, Twitter/X, GitHub, explorer) and a curated list of trusted crypto research voices, so users can verify findings themselves rather than blindly trusting a number.

How we built it

Built entirely with MeDo (Baidu's AI-powered no-code platform) using a fast, iterative "vibe-coding" workflow — ship, test, fix, repeat. The stack is React + Tailwind CSS with a terminal-aesthetic dark UI, pulling live data from CoinGecko's free public API with localStorage caching to gracefully handle rate limits.

The trust scoring algorithm runs entirely client-side, computing scores directly from real CoinGecko data fields — no external AI API calls, no rate limits, no latency.

Novus by Pendo is fully instrumented with 19 custom track events across all product areas, including trust_score_generated, trust_score_search_performed, and watchlist_item_added, ready to capture real user behavior as traffic scales.

Challenges we ran into

The biggest challenge was handling CoinGecko's free-tier rate limits without degrading the user experience — solved with a 5-minute localStorage cache and graceful fallback messaging. I also hit a platform constraint connecting Novus tracking to my deployed app domain, which I worked around using a GitHub Pages redirect

Built With

Share this project:

Updates