Inspiration
World changing to Automatic made me inspire to invent an automatic tool for analyzing and categorizing the failure of a job.
What it does
It analyzes and categorizes the failures, errors and warnings of a failed build pipeline automatically and also suggest the solutions based on the prediction from the historic data.
How we built it
A tool is built using a python script for collecting logs and using regex patterns it separates into failures, errors and warnings. Its analysis and categories based on the infrastructure. Using Machine learning it also suggest the solutions and in turn trains the model for better performance. Whenever a job gets failed, automatically this python script gets triggered. Also stores the results into the database.
Challenges we ran into
Requires great knowledge on a machine learning. Requires historic data for training the model.
What we learned
Became advanced in python programming Enhanced skills in Machine learning Developing a tool from scratch thought what all are the circumstances for the developer and get to rectify those.
Log in or sign up for Devpost to join the conversation.