Jovan Djukic

I work on a system called CocosAI that runs AI/ML workloads on combined datasets from multiple organizations while guaranteeing the privacy and security of the data and the algorithms by using Trusted Execution Environments. My focus was on the process of attestation and the secure boot feature.

I worked in a team whose main focus was reducing power consumption and increasing performance metrics of NVIDIA GPUs.

I participated in the following projects: - State Aptitude Test Grader (01/2021 – 12/2023) - the aim is to build a system that will automatically grade aptitude tests that students take at the end of a high school education. My team was tasked with building a desktop application using Python that will communicate with the scanning equipment and send scanned files to the server for further processing. - SMATSA (01/2020 – 07/2023) – the goal was to build a web application that will predict air traffic volume and enable air traffic controllers to do their job more efficiently. I was tasked with building the frontend using React framework. - UNDP sponsored prequalification to the IT sector (02/2019 – 09/2019, 09/2020 – 09/2021, 09/2022 – 07/2023) – the aim was to educate already employed people so that they are able to join the IT sector. I was teaching a course that covers Django web framework.

I worked at the Data Intensive Applications and Systems (DIAS) laboratory doing research on the (then novel) Knights Landing Intel microarchitecture.