The following is the information about AWS new services which is important for you to pass AWS exam. Amazon Kendra rebuilds enterprise search, which makes use of natural language processing and other machine learning technologies to unify multiple data islands within the enterprise, and consistently provides high-quality results for common queries, instead of a random list of linksresponding to the key word search. Amazon CodeGuru can help software developers automatically perform code auditand determine the line of code in the application with the most cost.
Amazon Fraud Detector assists companies to identify online identity fraud and payment fraud in real time based on the same technology developed for Amazon.com. Amazon Transcribe Medical transcribes audio messages into text with high precision in real time. This function is provided for healthcare providers so that they can concentrate on patient care. Amazon Augmented Artificial Intelligence (A2I) helps machine learning developers to check machine learning predictions through manual check. These are the five new services which will likely to be key knowledge in online aws testing.
Today, at the AWS reinvent Global Conference, Amazon Web Services, Inc. (AWS), a subsidiary of Amazon (NASDAQ: AMZN), released five new artificial intelligence (AI) services with a target at letting more application developers and end users to utilize machine learning. There is no need for them to have machine learning experience. AWS introduces several new services that apply AI so that more developers are able to apply machine learning to create a better end-user experience, including machine learning-driven enterprise search, code audit and analysis, fraud detection, medical transcription, and AI prediction manual audit. Machine learning continues to develop at a rapid speed. Nowadays, tens of thousands of customers are doing machine learning on AWS, including many customers who choose to use AWS entirely managed AI services, such as Alfresco, Bayer Crop Science, Cerner, CJ Cox Automotive, C -SPAN, Deloitte, Domino, Emirates NBD, Fred Hutchinson Cancer Research Center, FICO, FINRA, Gallup, Kelley Blue Book, Kia, Mainichi Newspapers, NASA, PricewaterhouseCoopers, White House Historical Society, Yamaha Corporation and Zola. Knowing this information will help you have a general understanding of the importance of AWS certification spoto exam. Over the past year, AWS has launched a number of new entirely managed AI services, such as Amazon Personalize and Amazon Forecast, which allow customers to benefit from the machine learning personalized recommendations and prediction technology. The technology is also used by Amazon consumers and highly praised by them. AWS customers are interested in learning from Amazon’s rich experience in using machine learning on a large scale to improve operations and offering a better customer experience while they don’t need to attend any training. They can also do optimization, and deploy their own customized machine learning models with more convenience. Today, AWS establishes five new AI services These services are based on Amazon’s abundant machine learning experience, allowing organizations of all industries and sizes to adopt machine learning in their enterprises without extra requirements for machine learning experience.
Amazon Kendra rebuilds enterprise search through machine learning
Though many efforts are made over the past years, internal search is still a trouble for many enterprises. Most employees still find it hard to look for the information they need. Institutions have a large amount of unstructured text data, which is very useful if it can be discovered and used to store multiple formats across different data sources (For instance, Sharepoint, Intranet, Amazon S3, and local file storage systems). Even though general-purpose, web-based search tools are ubiquitous, organizations still have trouble in carrying out internal search because there are no available tools that can make an index across existing data islands, failing to provide natural language queries and present accurate results. When employees have doubts, they need to use keywords that may appear in different contexts and multiple documents. These searches will usually generate a long list of random links. Employees must filter these links to find the information they need. Amazon Kendra enables employees to use questions (not limited to keywords) to search in multiple data islands, to deploy AI technology in the background so that the precise answers can be found (rather than a list of random links). The process works well in reshaping enterprise search. Employees can use natural language to search (keywords are still effective, but most users prefer natural language searches). For example, employees can ask a specific question, such as “When is the IT service desk open?” Amazon Kendra will give them a specific answer, such as “IT service desk opens at 9:30 in the morning”, and give directions to the IT portal and other relevant website links. Customers can use Amazon Kendra in applications, portals, and wikis. What they need to do is doing a few clicks in the AWS Management Console. Amazon Kendra can point to various document storage by customers. The service will aggregate petabytes of data to construct a centralized index. Amazon Kendra scans the authority of documents to make sure that the search results consist with the existing document access policies, and the search results only contain the documents that the user is permitted to access. Besides, Amazon Kendra actively retrains machine learning models according to customers’ specific situations, takes advantages of click rate data, user location, and feedback to improve accuracy. As time goes by, better solutions can be provided. Amazon CodeGuru uses machine learning to review automatic code so as to help organizations find the lines of code with highest expense, thus improving software development. The same as Amazon, AWS customers also write a large amount of codes. Software development is a well-known process. Developers write code, view the code, compile the code and deploy the application, measure the performance of the application and use the data to improve the code. If the code is incorrect at the very beginning, all the steps are useless.So the team will perform code checks including its logic, syntax, and format before adding new code to the existing application code base. Even for a large organization like Amazon, it is a tough task to recruit so many experienced developers having enough time to review such an amount of code every day. Even sophisticated auditors will miss some problems when encountering customer-oriented applications. This will result in errors and performance mistakes.
Amazon CodeGuru is a new machine learning service, automatically perform code audit and find the line of code costing the most in applications. It has two modules: code audit and application analysis. When it comes to code review, developers can submit the code as usual (at present support GitHub and CodeCommit. In the future, more stores will be supported), add Amazon CodeGuru to one of the code reviewers. They don’t need to modify already installed applications. When Amazon CodeGuru receives requests, it will automatically assess the code by using models that have gone through previous training. These models have been through code audit training in Amazon and GitHub for more than ten years. Amazon CodeGuru will check the modified quality, if any problem is discovered, it will require for adding legible annotations and mark out code line, specific questions and revise suggestions as well as example links and links pointing to related documents. Amazon CodeGuru also contains a application detector that is driven by machine learning, which can help customers find the most costing line of code. By using it,customers only need to install an agent in applications, and then Amazon CodeGuru can observe the running of applications and analyze application codes every five minutes. The code profile includes detailed information about latency and CPU utilization, directly linked to specific lines of code. Amazon CodeGuru help customers find the most costing line of code and create the flame graph, visually marking out other lines of code that cause performance bottlenecks. For years, Amazon internal team has used Amazon CodeGuru to analyze more than 80000 applications. From 2017 to 2018, the widespread use of Amazon CodeGuru internal version increases application efficiency and CPU utilization which is raised by 325% In terms of Amazon Prime Day team responsible for Amazon consumer business. The number of instances required to manage Prime Day has been reduced, and the overall cost has been reduced by 39%.
Amazon Fraud Detector detect automatic fraud by machine learning
Every year, organizations worldwide lose millions of dollars resulting from fraud. Nowadays, many AWS customers invest in large-scale and expensive fraud management systems. These systems are based on manual coding rules, thus consuming lots of time and money. They are also hard to catch up with the changes of fraud modes, leading to accuracy of systems lower than expectation. This also leads to organizations mistake excellent customers for fraudsters. They carry out more expensive fraud audit while missing the chance of lowering fraud rate. Over the past twenty years, Amazon is continually using top-notched technologies including machine learning to detect fraud knowing that it is a battle against fraudsters. To win the battle requires a large amount of resources to defense and keep updated with times. AWS customers hope that AWS can share professional knowledge and experience. Amazon Fraud Detector, based on the same technology used by Amazon consumer business, offers entirely managed service to test potential online identity fraud and payment fraud in real time without machine learning experience requirement. Amazon Fraud Detector makes use of the historical data of frauds and legal transactions to construct, train and deploy machine learning models and provide real-time fraud risk prediction with low latency. Firstly, customers upload transaction data to Amazon S3 and customize training models. Customers only need to provide e-mail address and IP relevant to transactions. They can choose whether or not to add other information such as billing address or telephone number. According to the type of fraud that customers want to predict (new account or online payment fraud), Amazon Fraud Detector will preprocess data, choose one algorithm and train a model–based on the ten more years of experience of Amazon running fraud risk detection and analysis on a large scale. Amazon Fraud Detector can also use machine learning based on Amazon-trained data detector. These detectors can recognize such fraud activities as abnormal regulation of e-mail naming on Amazon. Despite a small number of instances provided by customers to Amazon Fraud Detector, the accuracy of model training will also be improved. Amazon Fraud Detector can train models and deploy them onto entirely managed private API ends. Customers can send new activities like registration or new procurement to API and receive a fraud report including risk assessment. On the basis of the report, applications can affirm the right order like accepting purchasing or transmitting to manual auditing. With the help of Amazon Fraud Detector, clients can faster detect fraud with more ease and accuracy. To know more about Amazon Fraud Detector, please access: http://aws.amazon.com/fraud-detector.
Amazon Transcribe Medical use machine learning to conduct medical audio transcription so that medical care provider can focus on patient care. These days, part of regular work of doctors is to input detailed data into HER system. However, the approach they take to record details and patients’ conditions is poor. In many hospitals, doctors have to dictate medical notes to recorders. These audio files will be handed to third-party to do the manual transcription, therefore costing high service expense with long working time up to three days and filing progress delay. Another option is that utilize current advanced dictation software, but limited to current tools, doctors still haveto work for hours on clinical record. The third approach is that medical care providers employ transcribers who assist recording when doctors treat patients. Nevertheless, the existence of transcribers may make patients feel anxious. Their recordings also have some shortcomings and medical institutions find it hard to arrange and coordinate transcribers on a large scale. In short, the current solutions are insufficient in improving clinical recording efficiency and patient care. Amazon Transcribe Medical can solve all these problems through machine learning technology to conduct medical audio transcription automatically. Medical history application built on this function of Amazon Transcribe Medical can accurately and cost-effectively generate records. Amazon Transcribe Medical consists of many machine learning models which have gone through tens of thousands of hours medical audio training so as to provide precise and machine-driven medical transcription. The records formed in real time can save a lot of time. Besides, Amazon Transcribe Medical can also automatically transcribe conversations between doctors and patients. Without distracting attentions to write down records, medical providers can concentrate on patient care. Doctors can talk naturally. Amazon Transcribe Medical will use built-in automatic punctuation marks to break through the limitations of current transcription software. As for medical care providers, audio solutions built on Amazon Transcribe Medical can be applied to thousands of medical centers, thus eliminating the operative problem of managing and coordinating transcribers. Amazon Transcribe Medical satisfies HIPAA certification, provides user-friendly API and is capable of integrating with any equipment with audio applications and microphones. The export text of Amazon Transcribe Medical can also be used on other AWS service. For example, Amazon Comprehend Medical available of processing natural language can perform data analysis before finally going into medical history system. Amazon Augmented Artificial Intelligence (A2I) allows developers to use manual auditto check machine learning predictions. Machine learning can offer highly accurate predictions in various application scenes including identifying targets of images, extract texts from scanned documents or transcribe or understand spoken language. In every circumstance, the machine learning model will provide a prediction and a reliability score to present the certainty of the model’s prediction. The higher the score is, the higher the credibility of the result. When it comes to many user scenarios, when developers receive high-reliability results, they can trust that the results will be probably accurate and can automatically process them (for example, automatically adjust user-generated content on social networks, or add subtitles to videos). However, when the reliability is lower than expectation and the predicted results are inaccurate, manual check may be required to solve the ambiguity. This interaction between machine learning and manual audit is crucial to the success of the machine learning system, but the large-scale construction and operation of manual audit is very challenging with high cost which usually involves multiple steps. During these processes, customized software and a great number of auditors are needed to manage manual audit tasks and results. Consequently, developers often spend a lot of time managing the auditing process instead of building their intended application, or they have to give up manual check, leading to diffidence and low utility-rate of many predicted results.
Amazon Augmented Artificial Intelligence (A2I) is a new service. It can easily build and manage manual audit of machine learning applications. Amazon A2I offers service to common machine learning tasks including targets identification on images, audio language transcription and content review. With Amazon A2I, pre-built work flow of audit is provided, making it convenient to audit machine learning predictions of Amazon Rekognition and Amazon Textract. Developers select a believable threshold for a specific application, and all predictions with reliability scores below the threshold will be automatically sent to human reviewers for verification. Developers can choose 500,000 global labors from Amazon Mechanical Turk and pre-authorize labors’ third-party organizations such as Startek, iVision, CapeStart, Cogito, and iMerit, or their own auditors to perform audits. The checking results are stored in Amazon S3, and developers will receive notifications after the audit is completed so that they can take the next step based on the trusted results of the auditors. Amazon A2I brings manual audits to all developers and reduce the heavy workload of building and managing a customized review process or recruiting a large number of reviewers. Swami Sivasubramanian, vice president of machine learning at Amazon, said, “Companies in different industries have told us that they want to utilize Amazon’s rich machine learning experience to deal with some common challenges faced by companies. These challenges include bettering internal search and helping software developers to write better code, identifying fraud, and improve the overall quality of all machine learning systems. Amazon has also built an internal system that can successfully cope with these challenges with over ten years experience in building machine learning systems. Today’s release is a new round of iteration which motivates us to develop these systems. Through the release today, we are very happy that enterprises can use these machine learning functions without the need to have any machine learning knowledge.”
There are some comments from various companies on AWS’s new services as follows, knowing about it is helpful for you to understand the value of AWS certification.
3M is a multinational corporation, a leading manufacturers in such products as abrasives, chemicals, advanced materials, membranes, filtration, and adhesives. 3M applies technology to improve life quality through coordination. David Frazee, Technical Director of 3M Enterprise Research System Laboratory, said: “Research and development is the core of 3M. Science makes us stronger. This information is well kept in our patents and abundant knowledge storage. To look for proper information always makes people exhausted and time consumed. Sometimes, the information found is incomplete. By Amazon Kendra, our scientists can use natural language to search for information and accurately find the needed information. With the help of Amazon Kendra, Our engineers and researchers are enthusiastic to find information quickly, accelerate innovation, collaborate more effectively, and continually provide customers with unique products.” Work grid Software company is a wholly-owned subsidiary of Liberty Mutual with the purpose of providing software solutions for employees and making work more connected with higher efficiency and productivity. One of our core products is WorkgridChatbot. It enables employees to get the answers of frequently searched questions by working on a friendly and automatic natural language interface. A key part of the enterprise chatbots is to answer numerable questions from employees, so Workgrid provides a self-help quiz builder. The authors of content do not need to master programming language. They can train chatbots to reply to employees’ questions. Apart from these carefully planned content, we hope to provide WorkgridChatbot with an approach to easily extract knowledge from documents (such as PDF documents) across the enterprise,” said Gillian McCann, Head of Cloud Engineering and AI at Workgrid.” With Amazon Kendra, I am very glad that our customers can get the answers they need quickly and efficiently. Amazon Kendra is able to get answers directly from unstructured data in multiple repositories and track learning quickly, which enables us to deliver accurate and consistently optimized answers to our customers. We are happy to explore this function based on the combination of Amazon Kendra’s contextual intelligent search and task automation. Our employees can have a wonderful experience. ”
The BBC is one of the world leaders in the broadcasting industry. BBC videos can been seen all over the world. ”As a global media organization, we manage petabytes of video and broadcast live 24 hours a day,” said Matthew Postgate, Chief Technology and Product Officer of the BBC. “Amazon CodeGuru and other development tools used by our team are conducive to providing such a strong and reliable service to our audience continually and finding potential problems. It will also help us understand how to interact with the AWS platform, making the team reshape and optimize its codes so that people can get satisfactory service from BBC.”
Apptio SaaS solutions can make wise decisions for organizations when analyzing, planning and optimizing investments and transform IT operating models. Scott Chancellor, Chief Product Officer of Apptio, said: “Providing highly available and bug-free services to our customers is important to our success. We are always looking for tools to transform our organization, to more proactively detect problems at all stages of the application development cycle, and to improve development with less time spent on solving difficulties like concurrency, resource leaks, and performance bottlenecks. We have adopted Amazon CodeGuru and found that it can provide suggestions for solving these problems in the early stages of development. Furthermore, it can figure out the code area with slow service speeds. Hence, we can spend less time solving performance-related problems. These improvements will help us provide better experience for all customers.”
SmugMug+Flickr is one of the world’s most influential platform with photographer-center philosophy. SmugMug+Flickr is designed for professional photographers and amateur photographers where they can display their works and appreciate other people’s works. From the day of its founding, SmugMug’s motivation is to work out ideas letting photographers tell their stories in their own ways. When we expand our business, the functions of image processing, classification, and search are of key importance. SmugMug+Flickr CEO and chief geek Don MacAskill said: “Amazon CodeGuru’s real-time analysis does good to troubleshooting and identifying the inefficient parts of our services, especially the problem that the valuable lines of code in the application will slow down the speed of services. Amazon CodeGuru give advice, assists us to make modifications and optimizations. According to its suggestions, we are able to rebuild the code to make it highly maintainable and improve the performance of our services.”
Charles Schwab is an advanced investment service company. “Detecting fraud online is an endless challenge. The people with malicious intentions continually make attacks on media. Our mission is to protect our customers,” said Kara H. Suro, vice president of fraud monitoring and investigation of Charles Schwab. He added: “We are excited to see the establishment of Amazon Fraud Detector. Using it ,we can build machine learning tools faster and easier and detect fraudulent activities. The fraud prevention rates will also be increased significantly. Deploying Amazon Fraud Detector helps to identify fraud patterns from our historical data. We can also utilize Amazon’s experience in detecting fraud.
Vacasa is the largest full-service vacation rental management company in North America, supplying more than 23,000 vacation houses in 17 countries/regions. Every year, it serves more than 2 million guests. Eric Breon, founder and CEO of Vacasa, said: “Since the founding of the company, we have used technology to enable local teams to concentrate on caring for families and guests while maximizing revenue for vacation house owners. We are excited about the release of Amazon Fraud Detector. This entails that we can use advanced machine learning technology with more ease to accurately detect fraudulent bookings. To protect our house owners from potential risks. We can also put a focus on offering perfect vacation rental experience.”
Cerner is one of the leading providers of technology solutions, services and equipment of health information. Cerner’s solution strategist Jacob Geers said: “The accuracy of clinical documents is significant to the workflow and overall satisfaction of caregivers. Through Amazon Transcribe Medical’s transcription API, Cerner is developing a digital voice transcript gadget at incipient stage. It will automatically listen to the conversations between doctors and patients which are recorded in texts. Then, concepts will be translated intelligently and transmitted into Cerner’s medical history system. ”
Suki isa AI driven and audio-supported digital assistant. It can mitigate the burden of doctors. The CEO of Suki AI company said: “Clinical documents are related to the medical data workflow. The key lies in helping clinicians to take notes more effectively. We can easily integrate our clinical digital assistants with Amazon Transcribe Medical, allowing doctors to dictate medical notes, and efforts made on clinical documents can be reduced by 76%. Doctors should spend time caring for patients rather than importing data.”
As an Un-carrier in the United States, T-Mobile is redefining the way consumers and enterprises purchase wireless services through high-end product and service innovation. “At T-Mobile, happiness of our customers defines our success. As an Un-carrier, we know that customers will feel happy when we understand and predict their needs and directly solve their problems,” T -Mobile executive vice president and Chief Information Officer Cody Sanford said, “Our customer service model provided by our expert team is dedicated to establishing personal connections and using cutting-edge tools such as A2I to prepare our team for success. Indeed, machine learning can build up deeper and more engaged relationships. It realizes the access of real-time contextual information like detailed information of customers and available discounts. In this case, we can benefit from win-win results. We can conduct a real-time conversation with clients and represent them to make decisions.”
VidMob is a marketing creative platform that provides end-to-end technical solutions for all creative needs of brands. Its integrated platform combines first-of-a-kind creative analysis with top-notched creative making to improve marketing efficiency. “Vidmob uses machine learning to analyze various respects of videos, including people, objects, and information, to help brands better understand creative performance and work out better ideas. While for the aspects that are not covered by the existing machine learning models, we need to review ideas in petabytes of data that we analyze every day, which is quite challenging,” said Joline McGoldrick, senior vice president of data and insights at VidMob. “With our currently well-trained creative evaluation team, we can optimize and adjust accurately our prediction models at a faster speed by using A2I. We get in contact with many reviewers and the speed of models going on market is increased by three times.”
This information on AWS machine learning new services will probably appear in the AWS exam. Learning about it can be conducive to passing AWS exam. Relevant AWS exam questions you can find on SPOTO’s official website.