[ad_1]
Applied AI research to help Google Cloud enhance core solutions used by customers at scale
Alphabet’s Google Cloud helps organizations digitally transform themselves into smarter businesses. Its diverse solutions include cloud computing, data analytics, and the latest artificial intelligence (AI) and machine learning tools. Last week, we shared many of the platform’s latest advances at Next ’22, Google Cloud’s annual developer and tech conference on digital transformation in the cloud.
Over the past few years, we have partnered with Google Cloud to apply AI research to positively impact the core solutions our customers use. Here are some of these projects, including optimizing documentation comprehension, increasing the value of wind energy, and providing AlphaFold ease of use.
Supporting product innovation across Document AI
From cuneiform tablets to printing presses, society has created ways to share written knowledge quickly and effectively. These documents vary by country, language, and industry, making it difficult to find a solution for extracting and using such diverse information at scale.
Google Cloud’s Document AI enables users to simply upload documents such as invoices, tax returns, and bank statements to extract and query digital, printed, or handwritten scan information. Any industry looking to use AI tools for document understanding needs vast amounts of training data to work well. However, this kind of data is often unavailable, incomplete, or lacking proper annotation, which hinders widespread adoption of AI.
Worked with Google Cloud’s Document AI team to help develop an innovative machine learning model that required 50% less training data1 We are parsing utility bills and purchase orders and are continuing our internal evaluation of methods on other document types. He also works on improving solution performance across languages using smaller datasets, helping customers across different industries and geographies to leverage the benefits of his Document AI. .
Increase the value of wind energy
As part of our mission to solve intelligence to advance science and benefit humanity, we apply AI research to power global sustainability initiatives. These efforts are consistent with Google’s commitment to sustainability. We also recently partnered with Google Cloud Professional Services to positively impact the wind energy sector and help build a carbon-free future for all.
Wind farms are becoming an increasingly important source of carbon-free electricity, but winds can fluctuate depending on the weather. Wind farm operators rely on energy production forecasts to balance supply and demand on the power grid. Also, if the operator can commit to selling a certain amount of electricity based on the next day’s forecast, they will get a better price.
We worked with Google Cloud to help develop a custom AI tool to more accurately predict wind power output and train it based on weather forecasts and customer historical wind turbine data. Additional models recommend a delivery commitment one day in advance to bring this energy into the grid.
ENGIE, a global energy and renewable energy supplier, is currently piloting the technology in Germany and, if successful, could roll it out across Europe. Making wind energy more economically attractive while improving reliability will encourage the use of renewable energy – a win-win for everyone.
Providing access to the biggest breakthroughs in Vertex AI
From designing to deploying a new machine learning model, the development process involves many stages and requires a suitable data infrastructure. To help data scientists and enterprises along the way, Google Cloud built Vertex AI. It’s a single platform that gives users access to machine learning tools at every step of the development lifecycle.
Now available in Vertex AI, shortly after we released to the world the breakthrough AlphaFold system that accurately predicts the 3D structure of proteins. Using AlphaFold with Vertex AI makes it easier for a scientist to effectively run his AlphaFold prediction workflow by tracking experiments, optimizing hardware selection, and managing this at scale.
Earlier this year, we expanded the AlphaFold protein structure database to include nearly every cataloged protein known to science. We have partnered with Google Cloud to host this massive database, enabling bulk downloads of over 200 million proteins, with billions of structures already downloaded. This resource is rapidly becoming an essential tool for the scientific community, ushering in a new wave of biological progress.
Interested in applying AI research to make a positive impact on the world? Check us out open roll.
[ad_2]
Source link
