Head of Development at ShareTheBoard. Jerzy began his career in the telecommunications at Samsung R&D, where he worked as an engineer, technological leader, and group manager. Later, he moved into the world of Big Data and Data Management, leading engineering teams at companies like Gemius and CrediSuisse. In his free time, he writes articles for travel magazines and books.
Do you want to create a real-time video processing application? That's excellent, but have you considered how to handle delays, video compression, and high infrastructure costs? At ShareTheBoard, we thoroughly analyzed these challenges, tested various approaches, and ultimately succeeded in developing a web application that allows you to share digitized data with remote meeting participants. In this session, we will discuss the biggest challenges we faced and describe our final architecture.
The journey of every developer eventually leads to a point where their latest feature appears to be sluggish. This marks the ideal moment for optimization, a process that can be time-consuming, prone to numerous errors, and, in the end, potentially slow down the entire program. We know something about it because in ShareTheBoard, a web application designed for image processing using neural networks., we encountered almost every optimization pitfall. Furthermore, we have finally achieved success, allowing us to rent about what an effective optimization process should look like, the obstacles you may encounter when implementing machine learning models into the web world, and how to squeeze out the last drops of performance from a web browser.
