Issue link: https://read.uberflip.com/i/1444470
Document #US43529118TM ©2020 IDC. www.idc.com | Page 17 IDC White Paper | Accelerate Machine Learning Development to Build Intelligent Applications Faster The need and desire for better (and simpler) tools, quicker time to market, and efficiency are key concerns in the market for intelligent, AI-enabled applications. Organizations need guidance about what types of tools and technologies can help them develop intelligent, AI-enabled applications. They also need to understand when, why, and how these applications will be most effective in their organizations. In addition, organizations need to measure the effectiveness of these applications to determine return on investment for future projects that will include deep learning. Finally, the AI platform-as-a-service market is already crowded and is becoming more competitive with every passing day. The need and desire for better (and simpler) tools, quicker time to market, and efficiency are key considerations in the market for intelligent, AI-enabled applications. There are numerous established and emerging vendors addressing and providing services and solutions within this space at a very wide range of capabilities. As such, Amazon Web Services faces the challenge of continuing as a leader in this market and will need to maintain an aggressive pace of engineering and innovation. Although AWS is productizing machine learning/ deep learning services as the foundation of its solutions, this approach is not new to this market. What is new is that managed services such as SageMaker and the AWS Deep Learning AMIs combine numerous deep learning tools, frameworks, and technologies into a single integrated platform that provides significant productivity enhancements for organizations and developers. AWS needs to keep providing this level of innovation and expertise in this emerging market. Given the current state of innovation in deep learning and machine learning, it is important that these types of services remain framework agnostic and continue to accommodate the latest and greatest techniques and algorithms into the development process. Conclusion Critical success factors related to machine learning/deep learning implementation are related to people, process, and technologies. Traditionally, emerging technical solutions require sharp and motivated developers that like to live on the cutting edge of technology. However, cloud vendors are finding ways to democratize the development and use of AI and deep learning technologies to promote wider use and deployment within enterprises. The key is to quickly develop successful models and products based on deep learning. Some factors that can assist with this are: » Quick start packages/development tools. Some vendors offer templates, sample data, and sample code to help jump-start developer productivity. With managed services such as Amazon SageMaker Studio, data scientists and developers (and even nondevelopers) can be even more productive than they could be with just templates and sample code.