Innovative fault prediction can play a crucial role in ensuring smooth and cost effective process management. In this project, we designed a solution for predicting failure events by leveraging modern bleeding edge AI techniques such as deep learning, meta-learning, Bayesian optimization etc. Using AI we were able to successfully predict process/plant faults from telemetric data. Furthermore, correctly predicting the probability of failure events in the near future can serve as an immense catalyst for revolutionizing preventive maintenance. Our AI based solution was able to correctly predict time to fault; we tested our solution on a number of process plants. Our solution, by leveraging meta-learning, was able to perform extremely.