Using open source technology such as Kubernetes/Kubeflow to build a data science environment and ML pipelines will be covered next, before moving on to building an enterprise ML architecture using Amazon Web Services (AWS). Some of the business problems I worked on had complex rules, such as identifying companies for comparable analysis for investment banking deals, or creating a master database for all the different companies' identifiers from the different data providers. Each comment has been labeled with a. The business benefits this delivers are faster decision-making on loan application reviews and approvals, lower processing costs, and a reduced impact on a company's financial statement due to loan defaults. Section 1: Solving Business Challenges with Machine Learning Solution Architecture, Chapter 1: Machine Learning and Machine Learning Solutions Architecture, Chapter 2: Business Use Cases for Machine Learning, ML use cases in healthcare and life sciences, Section 2: The Science, Tools, and Infrastructure Platform for Machine Learning, Chapter 4: Data Management for Machine Learning, Hands-on exercise data management for ML, Chapter 5: Open Source Machine Learning Libraries, Core features of open source machine learning libraries, Understanding the scikit-learn machine learning library, Understanding the Apache Spark ML machine learning library, Understanding the TensorFlow deep learning library, Hands-on exercise training a TensorFlow model, Understanding the PyTorch deep learning library, Hands-on exercise building and training a PyTorch model, Chapter 6: Kubernetes Container Orchestration Infrastructure Management, Hands-on creating a Kubernetes infrastructure on AWS, Section 3: Technical Architecture Design and Regulatory Considerations for Enterprise ML Platforms, Chapter 7: Open Source Machine Learning Platforms, Open source technologies for building ML platforms, Hands-on exercise building a data science architecture using open source technologies, Chapter 8: Building a Data Science Environment Using AWS ML Services, Data science environment architecture using SageMaker, Hands-on exercise building a data science environment using AWS services, Chapter 9: Building an Enterprise ML Architecture with AWS ML Services, Key requirements for an enterprise ML platform, Enterprise ML architecture pattern overview, Hands-on exercise building an MLOps pipeline on AWS, Training large-scale models with distributed training, Hands-on lab running distributed model training with PyTorch, Chapter 11: ML Governance, Bias, Explainability, and Privacy. A truck delivery company optimizes the delivery route of its fleet to determine the delivery sequence required to achieve the best rewards, such as the lowest cost or shortest time. Lastly, you now have an understanding of how ML solutions architecture fits into the ML life cycle.
The number of ML models trained and deployed by some companies has gone up to tens of thousands from a few dozen models just a couple of years ago. Solution architecture also creates documentation that will be used to keep the system up to date, along with a deployment diagram, software patches, and a software release version, and enforces the runbook to tackle frequent issues and business continuation processes. I did not have a tool to track the different experiment results, so I had to track what I have done manually. The microservice architecture addresses the need for changing requirements in an agile environment, where any solution changes need to be accommodated and deployed rapidly. The second concept is to integrate this prediction workflow into a business workflow application. You can think of supervised ML as learning by example. In the life cycle of product development, the most challenging phase is to establish the nature of the requirements, especially when multiple elements are competing to be addressed as high priority and are evolving rapidly. This also aided in the gaining of a high-level overview of different cloud computing models, such as IaaS, PaaS, and SaaS, and the cloud computing deployment models in the public, private, and hybrid cloud. Based on the pain points and the availability of data, you can come up with some hypotheses on potential ML solutions, such as a virtual assistant to handle common customer inquiries, audio to text transcription to allow the text analysis of transcribed text, and intent detection for product cross-sell and up-sell. It covers all aspects of a system, which includes, but is not limited to, system infrastructure, networking, security, compliance requirement, system operation, cost, and reliability. This new edition features additional chapters on disruptive technologies, such as Internet of Things (IoT), quantum computing, data engineering, and machine learning. By the end of this book, youll be able to design and build an ML platform to support common use cases and architecture patterns like a true professional. The preceding diagram highlights the following attributes of a good solution architecture: Now you have had a high-level overview of solution architecture and its benefits, Let's investigate more closely the everyday aspects of solution architecture. This is similar to how unsupervised ML works. In the next chapter, you will learn all about the solution architect role itselfthe different types of solution architect, the role's responsibilities with regards to solution architecture, and how these fit into an organizational structure and agile environment. As a result, the agent will adjust its future moves to maximize the rewards in the future states of the environment. Without a properly identified business problem and its value proposition and benefit, it would be challenging to get an ML project off the ground. When users interact with software applications, they interact with functional requirements directly. Neelanjali evangelizes and enables AWS customer and partners in AWS database, analytics, and machine learning services. Solution architecture ensures an end-to-end solution delivery and impacts the overall project life cycle. As the project size increases, the team becomes distributed globally. As an ML solutions architect, you have been asked to validate an ML approach for solvin.
A delay in one task can impact the project timeline and can result in the organization missing the market window to launch the product. Model accuracy will not be a good metric to evaluate the model performance for a fraud detection use case this is because if the number of frauds is small and the model predicts not-fraud all the time, the model accuracy could still be very high. There is a huge demand for skilled ML solutions architects in different industries, and this handbook will help you master the design patterns, architectural considerations, and the latest technology insights youll need to become one. One of the most frequent answers I always get is about data that is, data quality, data inventory, data accessibility, data governance, and data availability. In this step, you would need to develop a clear understanding of the business goals and define the business performance metrics that can be used to measure the success of the ML project.
The following figure shows the key steps in ML projects: In the next few sections, we will discuss each of these steps in greater detail. The dataset was also small and did not require a large infrastructure for model training. This is usually done using a holdout dataset, also known as a test dataset, to gauge how the model performs on unseen data. Today, when you have multiple resources all around the world, a specific technology should be chosen very carefully. Based on my years of experience working with companies of different sizes and in different industries, I see ML solutions architecture as an overarching discipline that helps connect the various pieces of an ML initiative covering everything from the business requirements to the technology. An ML solutions architect interacts with different business and technology partners, comes up with ML solutions for the business problems, and designs the technology platforms to run the ML solutions. Solution architecture has evolved with technological modernization. The hyper-growth in the AI/ML field has resulted in the creation of many new professional roles, such as MLOps engineering, ML product management, and ML software engineering across a range of industries. Here, you will get a high-level overview of the different types of cloud computing deployment models. Enterprises choose to distribute their workload between different public cloud vendors to get the most out of each cloud technology or provide options to their team depending on their skill set. Instead, we would rely on ML techniques to help explain the relative importance of different input features to understand what factors were most influential in the decision-making by the ML models.
This could involve simple Proof of Concept (POC) modeling to validate the available dataset and modeling approach, or technology POC using pre-built AI services, or testing of ML frameworks. Through these responses, you will know what a good move is versus a bad move in order to stay alive and increase your score. Over the years, I have worked on many real-world problems using ML solutions and encountered different challenges faced by the different industries during ML adoptions. Solution architecture in the cloud has become increasingly important these days and is becoming the "new normal" as more enterprises choose to migrate their workload to it. The solution architect analyzes the functional requirements and defines non-functional requirements in order to cover all aspects of the solutionand avoid any surprises. The first thing I started to work on was to learn the sport. David has an engineering degree from Cornell University. While ML has a wide scope of applications, it does not mean it can solve all business problems. By the end of this handbook, you'll have learned the techniques needed to create efficient architecture designs that meet your business requirements. You will learn more about the details in Chapter 5, Cloud Migration and Hybrid Cloud Architecture Design. Public cloud vendors provide infrastructure and facilitate an array of services in various areas, such as analytics, machine learning, blockchain, robotics, application development, email, security, monitoring, and alerting. Taking continuous feedback and adapting to it is the key to high-quality delivery, which should adhere to all the phases of solution design and development. They think of an out-of-the-box idea to save the project from unforeseen issues, such as those covered in disaster recovery, and will prepare a backup plan in the event that things do not work out with the main one. He currently leads an AI/ML solutions architecture team at AWS, where he helps global companies design and build AI/ML solutions in the AWS cloud. As there are multiple public cloud providers in the market, you may start seeing the trends of multi-cloud. In general, the more training pictures with variations you have looked at during the learning time, the more accurate you will likely be when you try to name flowers in the new pictures. A recommendation engine optimizes product recommendations through adjustments based on the feedback of the customers to different product recommendations. As an ML architect back then, I mostly needed data science skills and general cloud architecture knowledge to work on those projects. Many people treat ML as a threat to their job functions. Saurabh led the AWS global technical partnerships, set his team's vision and execution model, and nurtured multiple new strategic initiatives. There are always risks and uncertainties involved during solution implementation; it can become very tedious should a developer need to spend time on fixing a bug, for example. There are two main deployment concepts here. There is a saying that data is the new oil, and this is especially true for ML. You will learn more about different architecture patterns in great detail in Chapter 6, Solution Architecture Design Patterns. These providers offer an array of services, from computing, storage, networking, databases, and application development, to analytics and machine learning. ML is a form of AI that learns how to perform a task using different learning techniques, such as learning from examples using historical data or learning by trial and error. Solution architecture handles critical, non-functional requirements such as scalability, high availability, maintainability, performance, and security while keeping business requirements in mind. After much trial and error, you will eventually be a very good player of the game. An example of ML would be making credit decisions using an ML algorithm with access to historical credit decision data. The field of artificial intelligence (AI) and machine learning (ML) has had a long history. Let's take a look at each of these elements: The goal of the business workflow analysis is to identify inefficiencies in the workflows and determine if ML can be applied to help eliminate pain points, improve efficiency, or even create new revenue opportunities. I was given some sample data, but I had no idea what to do with it. Let's learn about the various benefits of solution architecture in detail. Through my experiences, another key challenge that many companies have shared is gaining cultural acceptance of ML-based solutions. Cloud-native architecture primarily focuses on achieving on-demand scale, distributed design, and replacing failed components rather than fixing them. A third model is the hybrid cloud, used by large enterprises who are moving their workload from on-premises to a cloud, where they still have a legacy application that cannot move to the cloud directly, or maybe they have a licensed application that needs to stay on-premisesor sometimes, due to compliance reasons, they need to secure data on-premises. Their lack of knowledge of ML makes them uncomfortable in adopting these new methods in their business workflow. Becoming a solutions architect requires a hands-on approach, and this edition of the Solutions Architect's Handbook brings exactly that. It also includes updated discussions on cloud-native architecture, blockchain data storage, and mainframe modernization with public cloud. The solution architect develops a proof of concept and prototype in order to evaluate various technology platforms, and then chooses the best strategy for solution implementation.