Transformational AI: Take Inspiration from Successful Implementers

For most enterprises, artificial intelligence efforts are no longer science projects or skunkworks distractions. The technology has matured and companies are finding real value in pragmatic use cases that generate actionable insights or unlock new revenue streams. Success is possible for companies that prepare, invest, and partner with AI-fluent experts.

The opportunities around AI have now expanded significantly. Businesses rely on AI for customer service; to improve clinical outcomes in healthcare scenarios; to better assess financial risk; and predict the need for maintenance on manufacturing equipment to improve uptime.

The keys to these successes are bold IT leadership and collaboration within an AI center of excellence that joins together the infrastructure, data science, and business teams who can bring these applications to life.

“Most IT leaders want to be seen as enablers of business transformation,” says Tony Paikeday, senior director of AI systems at NVIDIA. “AI is the most transformative force taking hold in enterprises. This is a unique place where an IT leader can be seen front and center in transforming a business through platform, infrastructure and process, instead of just being a cost center that reacts to problems.”

While each AI project is unique and different, it helps to look back at previous success stories to inspire companies looking to run similar projects. Here are a few examples of how AI has helped to transform organizations:

• Domino’s delivers more than 3 billion pizzas a year. They wanted to leverage massive amounts of data and use AI to improve operational efficiencies and customer experience. Using purpose-built AI systems, they were able to quickly train complicated models — taking into account variables such as how many managers and employees are working, the number and complexity of orders in the pipeline, and current traffic conditions — in less than an hour (it had previously taken three days). By iterating more data, they were able to boost accuracy of their model from 75% to 95% for predictions of when an order will be ready.
• Lockheed Martin transformed its AI infrastructure from traditional CPU-based systems to a GPU-accelerated infrastructure and improved accuracy of models that predict asset health, minimizing downtime of fleets. Using natural language processing (NLP), they can analyze millions of maintenance records and determine risk levels around parts or components. Costs have been significantly reduced, as 9 out of 10 records today are classified without human involvement and with 95% accuracy.
• The Milwaukee School of Engineering needed additional computational resources and an optimized software stack to meet growing AI workloads. Thanks to the adoption of best-in-class AI infrastructure, today 80% of the institution’s computer science team actively uses the cluster and faculty GPU usage has increased by 10x. Now, students don’t need to worry that the “cloud odometer” is always running and limiting experimentation.
• St. Jude Children’s Research Hospital needed to address the growing computing demands of their data scientists and researchers, as well as siloed computing infrastructure — both of which were leading to increased cost of redundancies. They developed an AI Center of Excellence incorporating purpose-built AI systems into their HPC cluster to speed development. This central computing hub today provides them with all the computing resources they need to develop models that enable faster reading of radiology studies, faster genomic analysis, and cutting-edge research.

• BMW receives almost 10,000 new car orders a day, with 100 different options per car and 2,100 possible combinations. Using purpose-built AI infrastructure to train deep neural networks in a simulated 3D virtual world, BMW uses AI-powered logistics robots across their factory — from transporting materials to organizing parts. By using AI to create highly customizable just-in-time manufacturing, they are able to produce a new car in their factory every 56 seconds.

These examples are just the tip of the iceberg in terms of how AI can digitally transform a business and provide efficiency, optimization, and new revenue streams. With the assistance of NVIDIA’s DGXperts — AI-fluent practitioners who provide guidance and expertise — companies can get even more ideas to help them write their AI story.

So, chances are the time is right for you to start yours.

Uncover how to transform using AI with NVIDIA DGX Systems, powered by DGX A100 Tensor core GPUs and AMD EPYC CPUs.