Professionals are putting AI to work to turn our most valuable resource — data — into new ways of doing business. With AI, we are no longer wrestling with data, but using it to recommend with confidence, accelerate research and discovery, and enrich interactions with customers on their terms. The purpose of AI systems is to augment human intelligence, and today, we are excited to announce the next step on our journey to make AI more accessible for everybody with IBM Watson Studio.
Watson Studio: accelerating value for enterprises with AI
Watson Studio accelerates the machine and deep learning workflows required to infuse AI into your business to drive innovation. It provides a suite of tools for data scientists, application developers and subject matter experts, allowing them to collaboratively connect to data, wrangle that data and use it to build, train and deploy models at scale. Successful AI projects require a combination of algorithms + data + team, and a very powerful compute infrastructure.
Until today, there was a gap between data experts and domain experts. Only highly technical professionals in IT could organize and make sense of the vast amounts of data. Only domain experts could successfully convert data into the rich knowledge needed by AI. But domain experts and IT professionals worked in silos, with different tools and no visibility to each others work. The result was AI that fell short in its promise to augment people’s expertise.
Watson Studio closes the gap with a unified experience to create new insights from knowledge contained in the data. Watson Studio enables multidisciplinary teams across the organization to collaborate. We are convinced, after working with clients around the world, that rich collaboration is key unlocking the full potential of AI.
Comprehensive set of tools for the end-to-end AI workflow
In Watson Studio, we provide a choice of tools for the full AI lifecycle including best of-breed open source and IBM tools. You can choose between code or no-code tools to build and train your own ML/DL models, or easily retrain and customize pre-trained Watson APIs. Use the rich capabilities and controls to fine tune your models and automate the feedback loop of your models so they become smarter over time and continually adapt to changing conditions.
Connect and prepare data
In order to apply AI, the first step of the workflow starts with connecting and accessing data. Data scientists spend up to 80% of their time finding and preparing data, and 57% of data scientists said that cleaning and organizing data is the least enjoyable part of their job. The problem isn’t just limited to data scientists. Business analysts face similar struggles obtaining the data they need to build reports — often having to wait weeks for their IT team to extract data from the source systems.
To address the issue, we provide integrated capability to refine and wrangle data with Data Refinery, a tool that makes fast, self-service data preparation a reality. Watson Studio comes with more than 35 data connectors to the most popular data sources, whether they are in the IBM Cloud, 3rd party Clouds or application, or on-premises.
Watson tools and pre-trained models
Once you are connected to data, the next step is to build and train models. Application developers can get started with best-in-class pre-trained Watson APIs , which are the most accurate in the industry. These models will understand sentiment, classify topics in text, identify personality insights or recognize objects in a photo. We provide access to well documented APIs with samples and code snippets in the most popular programming languages.
Not all companies have access to talent and resources to the most advanced machine and deep learning technologies. That’s why we are offering simple tools to help transfer knowledge to Watson with your own data. The first integrated tool is for the Watson Visual Recognition service. This tool allows you to train custom models with your own images, to suit your specific visual recognition needs.
Choice of frameworks and best-in-breed tools
The machine and deep learning landscape is constantly changing. In Watson Studio, you will find support for the most popular tools providing users choice to easily train, save, deploy and automate the retraining of those models. They come pre-installed and we manage the underlying infrastructure for you so you can focus on your projects.
Use Python, R, or Scala in Jupyter Notebooks. Notebooks are a popular environment to create and share documents that contain live code, equations, visualizations, and explanatory text. Uses include data cleaning and transformation, numerical simulation, statistical modeling, machine learning, and much more.
We teamed with RStudio to deliver the most widely used open-source R statistical computing environment. Watson Studio includes the RStudio flagship product, a popular integrated development environment (IDE) that makes it easy for anyone to analyze data with R.
Visual modeling tools to power the non-coder
Although open source code in Python and R is popular because of its low cost, flexibility, and power, the time required to properly create code and ensure that it is working correctly can sometimes be frustrating. Talking to many customers, we understood that not everyone is a programmer. For those visual modelers we provide two tools:
- SPSS Modeler: an intuitive interface that is easy for everyone to learn and use — from business users to data scientists. Uncover valuable insights quickly for rapid time-to-value just by using drag-and-drop capabilities. There are thousands of Modeler users out there and the best news for them is that they can import and bring their SPSS streams within the Watson Studio — they will work and render as expected!
- Neural Network Modeler: an intuitive drag-and-drop, no-code interface for designing neural network structures. It speeds up the design process by avoiding the need to write and debug code by hand. Neural networks can be exported in TensorFlow, Keras, PyTorch and Caffe as well as in JSON format for sharing within blogs and code posted to GitHub.
Flexible compute environments to train models at scale
Today’s advanced AI models have grown in complexity and may require terabytes of data to train. There is a lot of innovation happening around increasing processing capabilities with CPUs, GPUs, distributed training and more. Watson Studio provides the flexibility to select the hardware you want for your experiments and lets you train models with unparalleled speed. You can quickly scale up or down your compute resources and customize your package dependencies in the environments.
Each compute environments is dedicated for each project collaborator, so no more competing for resources. We provide package management capabilities— the popular
conda environment definitions are used for customization. The compute environments are project assets and can be shared and reused by all members of your team for reproducible research.
Because our compute environment is in the IBM Cloud, you pay only for what you use. Our infrastructure is elastic and can handle the biggest spikes of use with ease to support the most demanding projects.
Model lifecycle & management
Moving a model into production is typically a tough task, and deployment requires help from busy IT specialists. When a single deployment can take weeks, it’s no wonder that most data scientists prefer to hand over their latest model and move onto the next project, rather than persist with the drudgery of continually retraining and redeploying their existing models.
In Watson Studio, it is possible to automate the retraining of models, and to monitor how the performance of those models evolve over time. That’s what we call Continuous Learning, which is unique to our platform. Thresholds can be setup, and if the performance drops, the user will get alerts and notifications so the data scientist can act.
Models are dynamic assets that need to be updated periodically, which is why it is key to have version control to roll-back to previous versions when needed — accessible through APIs and UI.
Share your results quickly with dashboards
Dashboards in Watson Studio allows users to painlessly add end-to-end data visualization capabilities to your application so your users can easily drag and drop to quickly find valuable insight and create visualizations on their own.
Interactive dashboards produce visualizations directly from your data in real-time. Smart data analysis and visualization capabilities help users discover underlying patterns and meanings in their data. Data can be explored using filtering and navigation paths. Embed dashboards into your application’s context, keeping users engaged.
Integrated with Watson Knowledge Catalog
We have developed Watson Studio to support the most demanding Enterprise AI use cases. For that, we integrated an intelligent cataloging service that allows you to bring together and prepare analytic assets, including structured and unstructured data wherever they live (on-premises or in the cloud), to turbocharge your data science, machine learning and AI.
Unite all your information assets into a single metadata-rich catalog. Watson-powered AI recommendations suggest the best assets based on Watson’s understanding of relationships between assets, and how they’re being used and socialized among users.
Powerful, integrated data policy activation engine, ensuring your sensitive data is automatically protected as determined by your governance policies.
Giving you more AI power
Watson Studio gives your models the chance to do what they were always meant to do: learn. By continuously training your models against the latest data, you can ensure that they continue to reflect today’s business realities, giving your organization the insight it needs to make smarter decisions and seize competitive advantage.
Watson Studio is the result of close collaboration with IBM Research and other IBM Cloud teams. We hope you enjoy the product and we can’t wait to see what you build with it. Give it a try here: www.ibm.com/cloud/watson-studio