Pizza Marketplace readers learned Tuesday about Domino's data- and AI-based efforts to create a better customer experience by using the power of graphics processing units to turn massive amounts of data around the pizza-making process into a better customer experience. The brand made an equally substantial investment in the complex technology with California-based GPU maker, Nvidia, for a relatively simple reason: Domino's wanted to tell customers when their pizzas were ready.
And anyone with any experience in the pizza restaurant business is aware that when customers are hungrily waiting for their pies, the answer to that question takes precedence over just about everything else.s.
"With CPUs, it used to take three days for Domino's to train an algorithm to predict when orders would be ready. Now using Nvidia GPUs, they can train the same algorithm in less than an hour and continuously retrain it using more data every day. This has enabled them to improve their accuracy from the previous 75% prediction accuracy rate to 95% accuracy."
-Azita Martin
Nvidia puts some of technology's best and brightest to work on answering that question for Domino's, using the vast stores of computing strength in GPUs to come up with accurate answers for waiting customers. Now, in the second half of this report, pizza operators get a closer look at how restaurants can use AI-based tech to elevate the customer experience for their own brands.
 |
| Azita Martin, Nvidia (photo provided). |
And the best part is that our source for answers to often complex computing questions — Nvidia General Manager of AI for Retail Azita Martin — has a special knack for breaking down the subject into answers that are easily understandable by all of us who do not make our livings in the tech sector, but do need to know how tech can produce money-making and saving benefits for pizza restaurateurs.
Q: First, can you connect the dots here to let restaurant leaders know why in the world they should even care about GPUs when it comes to improving their brands through the use of artificial intelligence?
A: AI can leverage massive amounts of data to predict certain outcomes and provide recommendations to humans. AI has multiple use cases for restaurants. … AI applications are based on trained algorithms (models) that predict or provide recommendations to humans.
The value of these algorithms heavily relies on their accuracy and ability to predict correctly, and for that you need massive amounts of data to train the algorithms. GPUs enable data scientists to create far more accurate algorithms by training and re-training models faster in minutes and hours rather than days and weeks using CPUs.
For example, with CPUs, it used to take three days for Domino's to train an algorithm to predict when orders would be ready. Now using Nvidia GPUs, they can train the same algorithm in less than an hour and continuously retrain it using more data every day. This has enabled them to improve their accuracy from the previous 75% prediction accuracy rate to 95% accuracy.
Q: Nvidia, specifically as a company, is well-known in the gaming world and elsewhere for its GPUs and 3D graphics. How did AI use for restaurant business object evolve from something as seemingly unrelated as Nvidia's work in 3D?
A:Nvidia has three core businesses: 3D graphics, HPC (High Performance Computing) and AI. We have invested in research and joint partnerships with thousands of AI startups for the past seven years and have deep expertise in AI.
"AI is demonstrating business value across every industry. But, in particular, there are many AI use cases that are helping retailers and restaurant chains improve gross margins."
-Azita Martin
Our secret sauce, beyond our GPUs, are the software libraries we have developed that accelerate machine learning, video analytics, speech recognition for conversational AI, and robotics. That has resulted in the most innovative startups using our GPUs for developing much more accurate AI applications. As a result, we have an ecosystem of thousands of startups that all leverage our GPUs, and we're now helping leading companies reinvent their retail and restaurant businesses with AI.
Q: Why is AI trending now for restaurant chains?
A:AI is demonstrating business value across every industry. But, in particular, there are many AI use cases that are helping retailers and restaurant chains improve gross margins. They are anywhere from making restaurants more efficient, reducing queues, improving forecasting and supply chain, to takeout food delivery.
Q: What areas make the most sense for restaurant brand to be considering for AI-related implementations now, in your opinion?
A:Some examples of AI use cases include:
- Video analytics to reduce food waste. (Ikea has worked with Nvidia on a project like this, Martin said).
- Domino's is using machine learning to predict when pizza orders would be ready to ensure customer satisfaction.
- Food delivery using robots (Martin indicated that Postmates is currently using AI in this manner).
- Video analytics applications to reduce shrinkage from repeat offenders that go from one store to another.
- Video analytics for notifying management to open more cashiers, when lines get long.
- Machine learning to improve demand forecasting and supply chain optimization that make sure the right meat and produce ingredients are at the right stores at the right time.
Q: What's coming next in AI for restaurant chains?
A: More AI innovation is quickly gaining traction across robot deliveries, logistics optimizations (including last-mile delivery) and speech recognition for taking curbside orders. These implementations can deliver significant operational efficiencies to restaurant chains.
Intelligent stores that leverage automation and ensure greater customer satisfaction are also of great interest to many business leaders. And, conversational AI is of great interest to many business leaders who want to provide a connected, personalized customer service experience to customers.
"pizza" - Google News
January 15, 2020 at 06:00PM
https://ift.tt/37909Za
What pizza operators can learn from Domino's use of AI - Pizza Marketplace
"pizza" - Google News
https://ift.tt/34pEGd3
Shoes Man Tutorial
Pos News Update
Meme Update
Korean Entertainment News
Japan News Update
No comments:
Post a Comment