What is birdeye-data-sprint4?

birdeye-data-sprint4 is a Python pipeline that pulls cryptocurrency market data from multiple sources, cleans and combines it, and produces a ranked list of tokens with risk and momentum scores, all without installing any third-party libraries. The pipeline fetches data from CoinGecko, which supplies broad market data, and DEX Screener, which supplies decentralized exchange pairs, with optional support for the Birdeye BDS API if you supply a key. It normalizes all three sources into a common format, removes obviously bad data such as tokens showing implausibly large price swings, and then deduplicates entries where the same token appears in more than one source. When merging duplicate entries, it flags how well the sources agree with each other as a data confidence rating of high, medium, or low. After cleaning, it calculates several scores per token: a momentum score that combines short, medium, and long-term price changes weighted by trading volume, a volatility measure drawn from recent price history, a volume anomaly flag for unusually high trading activity, and a safety score that blends liquidity, price stability, and data confidence. These feed into an overall composite score used to rank tokens and surface what the project calls hidden gems, meaning mid to small tokens with strong underlying numbers that have not yet attracted much attention. Running the pipeline produces three output files: a JSON file with the full analysis, a CSV file with 23 columns of token data, and a self-contained interactive HTML dashboard with sortable tables and color coded confidence and safety indicators. The dashboard also shows a stats summary, a sentiment breakdown chart, top gainers and losers, hidden gems, and volume anomaly alerts. The project runs with standard Python, tested on Python 3.9 and above, and uses no external dependencies at all, relying only on the built-in urllib, json, csv, and statistics modules. It is best suited for someone who wants a self-contained way to explore crypto market data without setting up a database or installing a large stack of packages.