Building intelligent APIs - Andy Thurai, IBM

Editor's Notes

• #4: All of this has led us to a new era of computing: Cognitive Systems Era. This is significant in the history of computing – there have only been 2 previous ones. Tabulating systems era – the advent of the adding machine to tabulate the census Programmable Systems era – Fortran, Cobol, PL1, etc. Cognitive – Watson Cognitive and Ambient computing
• #5: Years ago, we used to solder/glue things together when we needed to make things work. The so-called programming was hardware based. Then came the revolution – you can program computers. Now, at IBM we are taking this to the new level. Instead of scripting, writing, why don’t we just talk to the computer. You converse with computers. It is scary. Though I know couple of people sitting next to me, talking to their monitors allday. But that doesn’t count.
• #9: Man vs Machine. 1997. It all started way back. For many decades our engineers were working on it. The first time they got to show case some of those abilities were as Deep blue in 1997 against the then champion Gary Kasparov. So the computer won a chess match with a brainac. It proved that it can outthink a human. A major trend started. One of our researchers was watching Jeopardy, and thought well this involved a lot of thinking – why not this?
• #10: Watson went against the 2 of the best humans in Jeopardy - Ken Jennings and Brad Rutter who were the undefeated champs. Ken Jennings – 74 consecutive game wins, Brad Rutter – Highest money winner in jeopardy – EVER. Touted as the Carbon vs Silicon. Outsource the complex or boring work. We all know the outcome! Watson was not connected to the Internet, a stand along machine. Machine Learning, Statistical Analysis, NLP. Understand the questions, clues, copared the possible answers, ranking it with confidence, and their accuracy. Responded within 3 seconds in general. More importantly if it didn’t know, or confident enough about something it didn’t say a word. In day 2 came the defining moment, Ken explained it later, He guessed and thinking the double jeopardy was in a specific category, so did Watson. Not only that along with Ken it guessed the double jeopardy was in a specific category and they both went for it. Luck favored Watson in Day 2, that it landed the double jeopardy question and answered it right. At that moment it was over.
• #11: This is where we are now. Let’s look to the future. The Watson Ecosystem was launched in late 2013. We will scale it in 2014 as we open the aperture to bring more partners onboard.
• #12: Even a simple analytics project has multiple steps and people. Projects require multiple steps in order facilitate and adhere to the policies or practices that are important to the business process. However its rarely straight forward and does not address the needs of rapid and agile data exploration and discovery.
• #13: Projects require multiple steps like accessing and obtaining data. Business units and individuals from IT and departmental analysts are needed to manage and govern data sources. Providing this data to Analysts takes time and resources. Validation of this data is critical to avoid errors and poor decision making before the data is reviewed and approved for its original intent. If the data is wrong then the reporting will be wrong and decisions will be inadequate.
• #16: Watson in healthcare 90% of worlds data is created in last 2 years. 80% of todays data is not structured. 1 Trillion connected devices produce 2.5 quintillion bytes of data/ day. (1o power 18). Billion – Trillion – Quadrillion –Quintillion. 83% of CIOs indicate the need for BI and analytics on this almost impossible. 700K new articles by National Library of Medicine estimated to be in 2010. Physicians limited to 3-5 hours of reading per week. Not enough time to read journals, developments, etc. Not told by pharma reps but by a trusted advisor. Understand the human communication. 80% of decision making data comes from unstructured data. Clinical Situation. Overwhelmed by unstructed data.
• #17: Introducing IBM Watson Analytics. Watson Analytics offers automated data preparation and intelligence, engaging storytelling and guided predictions so you can find answers and action in your data on your own. No matter what part of the organization you’re in. Watson Analytics is such a revolution - a revolution in the techniques themselves and a further revolution in the convergence of all these technologies together. The result will be widespread adoption of advanced analytics in the business world. IBM Watson Analytics delivers a unified analytics experience on the cloud and helps you focus on the drivers that matter most in your business. By automating the steps of data access and refinement, predictive analysis, and visual storytelling, Watson Analytics immediately identifies and explains hidden patterns and relationships to accelerate your understanding of why things happened and what's likely to happen. Because Watson Analytics features natural language dialogue, you can ask the right questions and get results in the familiar terms of your business. Just as the first spreadsheet made financial calculations easier for anyone with a PC, Watson Analytics opens up the world of advanced analytics to all business users on the cloud. Think ahead Watson Analytics helps you go where your ideas take you with guided and predictive analytics and recommendations. Pursuing ideas, answering questions, making predictions, and deciding what to do next should be as easy as using an app on your smartphone. And you shouldn’t need help from IT or expertise from data scientists. With Watson Analytics, making decisions, anticipating outcomes and taking action can now be a smooth and interesting process.  And, if you aren’t sure what you need to do, Watson Analytics automatically starts you off with ideas and guides you through analysis so you can find meaning in your information and make better decisions. It even introduces additional capabilities you might find useful as you become bolder and want to dig deeper.Make more informed decisions to move your business forward. Tell a story Visualization is the illustration of trends and patterns in your data so they can be understood. It is a critical capability of analytics. Having the right visualization can make the difference between seizing the right opportunity and making a sketchy decision. However, choosing just the right one can be tricky. Watson Analytics recommends the visualizations that illustrate what’s important the most effectively and brilliantly. You can use these visualizations to create infographics that advance understanding and communication and spur the right action.Ask questions and receive answers in the language you use in your business. Understand your business Watson Analytics jumpstarts your analysis. It automates intelligence to accelerate your ability to get to the answers you’re seeking. It immediately starts you off with a visual story that illustrates what you need to know. Instead of fumbling over data or searching for answers, you can focus on understanding your business and effectively communicating results to others. Simply type in what you would like to see and Watson Analytics produces comprehensive results that explain why things happened and what's likely to happen, all in familiar terms. And as you interact with the results, you can continuously fine-tune your questions to get to the heart of the matter.Tell a clear, insightful story through rich data visualization. Get better data Built into Watson Analytics are information management services for data access, refinement and management. Cloud-based data management services automatically find, acquire and improve data sets with the touch of a button. And, the latest in columnar and in-memory technology handles large volumes of data with ease. Watson Analytics can also score the readiness of your data for analysis, highlighting potential data issues that could compromise results. What does this mean for you? Basically, you don’t have to worry about the state of your data. Watson Analytics is designed to make sure the right data is ready when you are. With automation and embedded data refinery services, you get better data and better understanding. Whether you are in marketing, sales, IT, operations, HR or finance, you have the trusted data and intelligence you need to complete your projects. Take advantage of automation — from data preparation to analysis.
• #18: What are the features of Watson Analytics Cloud-based Agility Watson Analytics is available anytime, anywhere on the cloud. Self-service for Business Users Watson Analytics removes the reliance on DBAs, Statisticians, Report Authors or other specialists to do analysis. Its intuitive interface and unique automation of data access and refinement, advanced analytics and visual storytelling empowers business users to do more by themselves, like having a Data Scientist in a Box. Start anywhere, and let your curiosity take you wherever you want to go. Watson Analytics suggests where you might want to go next, progressively exposing you to additional capabilities you might want to leverage to gain a deeper understanding of your business.   Differentiator Business users want a consumer-like experience to use analytics and data to pursue ideas, get answers to all types of questions, make predictions, and communicate compelling recommendations without needing help from anyone. However, they are hampered by the usability of analytics tools, access to data and dependency on others – both IT to provide the infrastructure and data scientists to provide the analytical expertise. With Watson Analytics, data-driven decisions can now be an essential, daily and interesting activity for business people. Data Discovery Watson Analytics lets you visually explore and interact with your data to find the obvious patterns and derive useful insights. Mobile-ready Watson Analytics is optimized for tablet devices for analysis on-the-go. Natural Language Dialogue Watson Analytics speaks the language of your business. Simply type in what you would like to see and Watson Analytics produces comprehensive results that explain why things happened and what's likely to happen, all in the familiar terms of your business. And as you interact with the results, you can continuously fine-tune your questions to get to the heart of the matter.    Differentiator Other offerings expect you to adapt to the software – speaking in some form of programming language. Watson Analytics lets you carry on a natural, iterative dialog using the terms that are meaningful to your business.
• #19: What are the features of Watson Analytics Single analytics experience  Watson Analytics brings together a complete set of self-service analytics capabilities on the cloud. You bring your problem, and Watson Analytics helps you acquire the data, cleanse it, discover insights, predict outcomes, visualize results, create reports or dashboards, and collaborate with others.   Differentiator Unlike other offerings that make you decide what type of analytics you want to do, and then select separate tools to accomplish it, Watson Analytics is a seamless, unified experience that lets you go where your ideas take you. For example, you can move seamlessly from discovery to predictive analysis – without changing user interfaces or losing work. Fully automated intelligence Just bring your data, and Watson Analytics will do the rest. By automating all the steps of data preparation, predictive analytics, and visual storytelling, Watson Analytics jumpstarts your analysis and accelerates your time to value. It immediately starts you off with a visual story that illustrates what you need to know. Instead of fumbling over data or searching for answers, you can focus on understanding your business and effectively communicating results to stakeholders.   Differentiator  Most analytic offerings assume you have data ready for analysis, a clear idea of the type of analysis you want to do, and the skills and time to build a model for analysis. However, most business users have none of these things. Data prep and loading can represent 60% or more of the time in an analysis project. Business users can spend a lot of time figuring out what analysis would be relevant and how to tell the story in a report or diagram. Watson Analytics automates these steps to accelerate your ability to get to the answers you’re seeking. Guided Analytic Discovery Watson Analytics features the use of predictive analytics to surface the most relevant facts and uncover unforeseen patterns and relationships. This sparks the right questions to ask and directs your attention to the parts of their business that matter most.   Differentiator Most discovery solutions expect you to decide what’s important in your data for further analysis. Watson Analytics lets you explore as well, but often business users aren’t sure what type of analysis would be the most meaningful. Watson Analytics automatically surfaces what matters most in your business. You can then use that insight to guide your analysis.   Data Access & Refinement Watson Analytics automatically shapes and joins data and then uniquely scores the readiness of your data for analysis, highlighting potential data issues that could compromise results....(access to data, data shaping, data quality/cleansing, data masking) Integrated Social Business Watson Analytics help you to quickly and securely build consensus around your data-driven analytical findings. Overcome analysis paralysis by collaborating with all stakeholders to share, discuss, combine and build the evidence necessary to move forward on decisions. Reporting & Dashboarding Watson Analytics takes the guesswork out of how to best show what's important in your data and automatically recommends the most effective visualization to convey the results you are looking for. With a few clicks, users can further alter chart types and data to fully customize the interactive reports and dashboards to meet the needs of the business. Visual Storytelling Watson Analytics helps users to quickly consolidate and annotate key visualizations into powerful infographics to elegantly and effectively convey results to stakeholders.
• #20: About 100 years later, we are in a situation to do exactly that. We are in a situation to mine the data, dynamically update the predictive models with the stream of current and live data and be able to predict the future.
• #21: Learn more about IBM Watson Analytics at WatsonAnalytics.com. Visit the site to get started with Watson Analytics for free. Register to access Watson Analytics on the cloud. Watch demos and how to videos, read content and talk to our experts through our community forum. Four key takeaway points: Watson Analytics brings together a complete set of self-service analytics capabilities on the cloud. You bring your problem, and Watson Analytics helps you acquire the data, cleanse it, discover insights, predict outcomes, visualize results, create reports or dashboards, and collaborate with others. Just bring your data, and Watson Analytics will do the rest. By automating all the steps of data access and refinement, predictive analytics, and visual storytelling, Watson Analytics jumpstarts your analysis and accelerates your time to value. It immediately starts you off with a visual story that illustrates what you need to know. Instead of fumbling over data or searching for answers, you can focus on understanding your business and effectively communicating results to stakeholders. Watson Analytics speaks the language of your business. Simply type in what you would like to see and Watson Analytics produces comprehensive results that explain why things happened and what's likely to happen, all in the familiar terms of your business. And as you interact with the results, you can continuously fine-tune your questions to get to the heart of the matter. Watson Analytics features the use of predictive analytics to surface the most relevant facts and uncover unforeseen patterns and relationships. This sparks the right questions to ask and directs your attention to the parts of their business that matter most.
• #25: Please note: Presenter notes are on every slide. Written as a suggested script to help with messaging.
• #28: Key themes Speed Instant environments Quick deployments Sign up in minutes Ease of Use Instant Environments Services prebuilt for your use - IBM, Third Party, Community DevOps tools to monitor, plan, deploy, and manage your apps Flexibility On-Prem integration Flexible pricing Security IBM secures the platform and infrastructure - leveraging experience with softlayer and proven on-prem security implementations Provides you with the tools to secure your apps
• #29: While timing is critical (what we just discussed), the systems and data you use today can’t be forgotten. That’s why Bluemix simply and securely connects to core IT and even other clouds (leveraging IBM’s proven reputation as a systems integrator). Hybrid cloud, where Public and Private clouds work together with Core IT investments, is critical to success given the app revolution and the expectations and increased competition it has generated. Overall, hybrid cloud represents the “New IT…”
• #30: Bluemix is built on top of IBM’s infrastructure as a service offering - SoftLayer. It embraces Cloud Foundry as an open source platform as a service and extends it with IBM, third party, and community built services. Explain Diagram (verbs in blue - please use this terminology) Start by Pushing your code through Cloud Foundry Cloud Foundry interprets which language you’ve used and associates your code to the appropriate runtime - now you have a working app (layer above runtimes). Your app can Pull prebuilt services (IBM, 3rd party, or community built) in from the marketplace and utilize their functionality Connect to traditional IT - aka “Your Systems” with a secure connector Connect to other clouds that you utilize with a secure connector All of this sits on infrastructure (networking, storage, servers) hosted by IBM Softlayer (IBM’s IaaS cloud offering) Finally - when your app is built and ready to go, it can be accessed by anything with a web browser and a connection to the internet, a smartphone app (via the Bluemix SDK) to utilize backend services, or another type of app or system (via an API you create).
• #31: Bluemix allows developers the flexibility to choose their own languages and tooling - but choice doesn’t stop there. We want to allow developers to compose their applications of whichever services, IBM created or otherwise, that work best for them, their organization, and their projects. IBM Created Service Examples (full range of proven, industry leading capabilities): SQL Database (DB2 relational database), Analytics Warehouse (Blu Acceleration powered), MQLight (message provider), Cloudant (IBM’s no-sql database), Watson Services (cognitive computing), Mobile Application Security, Push Notifications Third Party Created Service Examples (competing or otherwise - note the colors): Clear DB (MANAGED MySQL relational database), Cloud AMQP (MANAGED rabbit MQ message provider), MongoLab (MANAGED no-sql mongo db), Square (process mobile payments), Twilio (VoIP Calls, SMS & MMS sending/receiving) Community Services (if open source works best for you): mysql (MySQL open source db), rabbitmq (RabbitMQ open source message provider), redis (open source Key-Value store),
• #32: More on the Watson services in Bluemix can be found here: www.ibm.com/smarterplanet/us/en/ibmwatson/developercloud/
• #34: We know how important your existing on-premises and other cloud investments are to your business - so we’ve designed Bluemix to quickly, easily, and securely connect to them: Secure Connector provides a simple mechanism for connecting to another system - be it on-site (behind a firewall) or in the cloud. If you have DataPower, Bluemix works with that as well. The Standard (Cast Iron) Secure Connector is a simple software based connector that establishes a tunnel between BlueMix applications and the network on which it is installed, leveraging a secure (HTTPS) access that eliminates the need for a firewall port. You can create REST APIs against endpoints you’ve securely connected to (i.e. an API that “GETs” customer information from a DB2 database on-premises) CastIron Live (separate IBM SaaS product) Integration Data Orchestration (future) Catalog of the custom APIs you’ve created that is in-line with existing service catalog (future)
• #37: User Modeling -  Personality profiling to help engage users on their own terms Question and Answer - Direct responses to users inquiries fueled by primary document sources Machine Translation- Direct responses to users inquiries fueled by primary document sources Message Resonance - Communicate with people with a style and words that suits them Concept Expansion - Maps euphemisms or colloquial terms to more commonly understood phrases Relationship Extraction - Intelligently finds relationships between sentences components (nouns, verbs, subjects, objects, etc.) Visualization Rendering -Graphical representations of data analysis for easier understanding Language Identification - Graphical representations of data analysis for easier understanding