Designing with AI is challenging due to its complexity and unpredictability. AI models can operate like black boxes, making it difficult to interpret or explain their decisions. Ensuring accuracy requires extensive data and careful model training. AI systems must also be optimized for performance, balancing trade-offs like accuracy, speed, and resource consumption. At the AI Day Virtual Conference, we will explore each of these avenues, looking at the biggest challenges, with potential solutions to each.
Thu, Sep. 25, 2025 · 8:00 am (PDT)
Executive Vice-President, Brand Director, Embedded Franchise, OpenSystems Media
Thu, Sep. 25, 2025 · 8:45 am (PDT)
View this session
A key engineering challenge when it comes to designing systems based on AI involves ensuring reliable, ethical, and trustworthy systems, as well as addressing scalability, data quality, and the need for specialized expertise. This webinar explores the critical engineering challenges in AI system design, focusing on building reliable, ethical, and trustworthy technologies. Topics include ensuring system scalability, managing data quality, and navigating the demand for specialized expertise. Attendees will gain insights into best practices, emerging standards, and practical strategies to overcome these obstacles, enabling the development of robust and responsible AI-driven solutions.
A key engineering challenge when it comes to designing systems based on AI involves ensuring reliable, ethical, and trustworthy systems, as well as addressing scalability, data quality, and the need for specialized expertise. This webinar explores the critical engineering challenges in AI system design, focusing on building reliable, ethical, and trustworthy technologies. Topics include ensuring system scalability, managing data quality, and navigating the demand for specialized expertise. Attendees will gain insights into best practices, emerging standards, and practical strategies to overcome these obstacles, enabling the development of robust and responsible AI-driven solutions.
Executive Vice-President, Brand Director, Embedded Franchise, OpenSystems Media
Thu, Sep. 25, 2025 · 10:00 am (PDT)
View this session
Large language models (LLMs) are quite common in today’s AI-based systems. However, these models can require substantial computational resources, including powerful GPUs and large memory, both for training and inference. Ensuring reliability can also be tricky, as LLMs require rigorous testing and monitoring. In this session, we will look at managing the LLMs with respect to size and complexity, and scalability, for integration into real-world applications. That includes balancing trade-offs between performance, cost, and user privacy.
Large language models (LLMs) are quite common in today’s AI-based systems. However, these models can require substantial computational resources, including powerful GPUs and large memory, both for training and inference. Ensuring reliability can also be tricky, as LLMs require rigorous testing and monitoring. In this session, we will look at managing the LLMs with respect to size and complexity, and scalability, for integration into real-world applications. That includes balancing trade-offs between performance, cost, and user privacy.
Executive Vice-President, Brand Director, Embedded Franchise, OpenSystems Media
