Abstract

Artificial intelligence is becoming a key technological differentiator and companies are tasked with taking the promise of AI and integrating it into their businesses. This is no trivial task, and AI solutions should not be limited to companies with large engineering teams. This talk will outline some key requirements and practical design of an AI production pipeline. It also will cover some practical examples of how Intel’s publicly available reference kits can simplify this process and significantly reduce time to market with the latest hardware and software optimizations.

Brief Biography

Tony Mongkolsmai is a software architect and technical evangelist at Intel. He has 20 years of industry experience designing, implementing, and leading production grade software solutions. He was a lead architect for Intel’s performance tools teams, which included Intel VTune Performance Analyzer. He also led an engineering team that designed and implemented a Cloud Native, scale out AI Data Center for Intel. He also was part of the group that successfully proposed the creation of a Sustainability Technical Advisory Group for the Cloud Native Computing Foundation (CNCF). Tony now helps drive software strategy and evangelizes for Intel in a variety of software spaces including Cloud, AI and oneAPI toolkits. You can hear Tony talk regularly about a variety of technology topics like AI, Quantum Computing and Cloud software design as host of the Code Together podcast. He likes to focus on open-source solutions and driving solutions that promote ease of use and rapid time to market.