Challenges to the credibility of Machine Learning pipeline output. How the performance of such ML models are inherently compromised due to current practices. How such problems can be cured by using the quality, improved access to data saved using smart-contracts and the blockchain.