The Institute will apply detection, prediction, and explanation techniques to develop an online machine learning tool capable of identifying and responding to errors and anomalies.
Managing Information Technology (IT) systems – complex and dynamic by nature – is no easy task. These systems, able to execute thousands of operations simultaneously, can generate an endless number of alerts; many of these do not require immediate attention, but some signal serious issues.
How can operators decide which “fires” to put out? And is there a way to anticipate them and reduce their impact? These are the questions that the OnlineAIOPS2 project, led by INESC TEC, aims to answer.
According to João Mendes Moreira, researcher at INESC TEC, over the next three years the team will develop “methods for real-time collection and processing of large volumes of events generated by infrastructures and applications.”
Based on this data, it will be possible to detect and diagnose anomalies, explain said behaviours, isolate relevant events, and recommend actions to operators. The goal? To provide the global market with an online machine learning framework capable of addressing the challenges of IT management in real time, while assisting operators in problem-solving procedures.
OnlineAIOPS2 builds on the work carried out by INESC TEC in partnership with the company IT PEERS, between 2021 and 2023. While that first phase advanced anomaly detection, the second phase aims to reach a higher Technology Readiness Level (TRL), further develop the results, and move towards commercialisation.
The development of detection, prediction, explanation, and recommendation techniques will be led by INESC TEC, with contributions from the Faculty of Engineering of the University of Porto and the Faculty of Economics of the University of Porto.
The project will run for three years and has been funded under the Portugal 2030 programme.
The researcher mentioned in this news piece is associated with INESC TEC and FEUP
News, current topics, curiosities and so much more about INESC TEC and its community!
