A World with Watson

An year back I wrote my first blog about Watson. I have been closely following what’s happening with Watson. Here are some facts on Watson and what user’s of Watson are speaking about it.

 

Quick Facts About Watson:

  • By the end of this year, Watson will touch one billion people in some way
  • Watson can “see,” able to describe the contents of an image. For example, Watson can identify melanoma from skin lesion images with 95 percent accuracy, according to research with Memorial Sloan Kettering.
  • Watson can “hear,” understanding speech including Japanese, Mandarin, Spanish, Portuguese, among others.
  • Watson can “read” 9 languages.
  • Watson can “feel” impulses from sensors in elevators, buildings, autos and even ball bearings.
  • Watson has been trained on 8 types of cancers, with plans to add 6 more this year.
  • Beyond oncology, Watson is in use by nearly half of the top 25 life sciences companies, major manufacturers for IoT applications, retail and financial services firms, and partners like GM, H&R Block and SalesForce.com.
  • At IBM, there are more than 1,000 researchers focused solely on artificial intelligence

But perhaps more important than what Watson can do, it is what people, businesses and institutions of all sizes are doing with Watson. See what some IBM Watson users are saying.
Watson
What IBM and Watson has been at the leading edge of is providing enterprise grade, commercially ready cognitive services, fully integrated with a top notch cloud and many other services from analytics to support and sales & marketing.”  — André M. König, Co-Founder @ Opentopic Inc. This quote was included in Mr. König’s article “Watson is a Joke?” featured on LinkedIn.

All of us involved in training Watson… are absolutely convinced that this technology will become an indispensable part of a doctor’s armamentarium to care for patients.” — Mark G. Kris, MD, lead physician of the Memorial Sloan-Kettering-IBM Watson collaboration. Dr. Kris’s quote was featured in a June 25, 2017 article in the American Society of Clinical Oncology entitled “How Watson for Oncology is Advancing Personalized Patient Care.”

But, the probably more exciting part about it is in 30 percent of patients Watson found something new. And so that’s 300-plus people where Watson identified a treatment that a well-meaning, hard-working group of physicians hadn’t found.” Dr. Norman “Ned” Sharpless, director of the Lineberger Comprehensive Cancer Center at the University of North Carolina at Chapel Hill and recent presidential appointee as director of the National Cancer Institute.
Dr. Sharpless’ made these comments in a “60 Minutes” segment that aired on October 2016 and again on June 25, 2017. The segment can be viewed here.

30 minutes is down to 8 minutes to screen a patient…That coordinator can now spend that valuable time gained … in educating the patient on why it’s important for her to be in that clinical trial, helping to break down other barriers.”  Dr. Tufia Haddad, MD, Breast Medical Oncologist, Mayo Clinic, made these comments during an AI in Healthcare panel during HIMSS 2017, reported here.

We could have individually looked at the 1,500 proteins and genes but it would have taken us much longer to do so.  IBM Watson for Drug Discovery, with its robust knowledge base, was able to rapidly give us new and novel information we would not otherwise have had.” – Robert Bowser, PhD, director of the Gregory W. Fulton ALS Research Center at Barrow Neurological Institute and one of the nation’s leading ALS researchers. Quote is from a press release announcing the recent Society for Neuroscience study findings.

[With Watson], we’re seeing some really tremendous efficiencies gained in the drilling business – [including] an 80 percent reduction in the geoscience research time we need to actually design our wells. That means geoscience searchers are doing geoscience not looking out for more data.” -Peter Coleman, CEO and Managing Director for Woodside [source:  Investor Briefing, March 7, 2017]

[Watson services] was a wake-up call for us – that cognitive solutions are real and powerful. We felt that IBM had, by far, the largest lead in terms of where cognitive was going and that the Watson team would be in the best position to help our business users.” -Ryan Bartley, Head of Applied Innovation at Staples [source: IBM Watson blog, February 10, 2017]

It’s not man versus machine—they very much work hand and hand. Our analysts continue to play a critical role in evaluating a cyber security incident, while Watson for Cyber Security enforces their decisions and validates what they are sharing with the customer at risk. It enables security analysts to deliver faster and more accurate details on a breach, so we may better protect our customers.” – Ronan Murphy, CEO, Smarttech (source: Press Release, May 11, 2017)

Why Blockchain?

There has been a lot of buzz on blockchain taking it to Gartners Hype Cycle for Emerging Technologies, 2016. It has been envisioned that blockchain will do for transactions what the Internet did for information. So in this blog, lets discuss the need for blockchain?

Why Blockchain?

MultipleLedgers
Complex Transactions

If you’ve ever bought a house, you probably had to sign a huge stack of papers from a variety of different stakeholders to make that transaction happen. It is a complex transaction involving banks, attorneys, title companies, insurers, regulators, tax agencies and inspectors. They all maintain separate records, and it’s costly to verify and record each step. That’s why the average closing takes several days. Same holds good if you are registering a vehicle. In these two examples, what you are doing is ‘Establishing ownership of the asset’ and the problem is that there are several ledgers (or databases) where the information resides and all of them have to have the same version of truth. So the problem are many fold:

  • Multiple ledger(s) which are updated to represent business transactions as they occur.
  • This is EXPENSIVE due to duplication of effort and intermediaries adding margin for services.
  • It is clearly INEFFICIENT, as the business conditions – the contract – is duplicated by every network participant and we need to rely on intermediaries through this paper laden process.
  • It is also VULNERABLE because if a central system (e.g. Bank) is compromised due to an incidents this affects the whole business network.  Incidents can include fraud, cyber attack or a simple mistake.

Solution:

What if there existed a common ledger (or a distrubuted database) that everyone had an access to and everyone trust? This is what blockchain does to the business!

Why now?

There are three reasons why blockchain is starting to take a foothold now.
  • Industries are merging and interacting like never before. The growth of ecommerce, online banking, and in-app purchases, and the increasing mobility of people around the world have fueled the growth of transaction volumes. And transaction volumes will explode with the rise of Internet of Things (IoT) — autonomous objects, such as refrigerators that buy groceries when supplies are running low and cars that deliver themselves to your door, stopping for fuel along the way. These partnerships require more trust and transparency to succeed.
  • There is increasing regulation, cybercrime and fraud that is inhibiting business growth. The last 10 years have seen the growth of global, cross-industry regulations, including HIPA, Sarbanes -Oxley Act, anti-money laundering and more. And to keep pace with regulatory changes, companies are rapidly increasing compliance staff and budgets.
  • Advancement in technologies like cloud (offering compute power to track billions of transactions) and cryptography (securing both networks and transactions) are also enablers for blockchain.

In my future blog I will discuss how blockchain makes things better and how it works. So stay tuned.

Data Science Vs BI & Predictive Analytics

Business intelligence (BI) has been evolving for decades as data has become cheaper, easier to access, and easier to share. BI analysts take historical data, perform queries, and summarize findings in static reports that often include charts. The outputs of business intelligence are “known knowns” that are manifested in stand-alone reports examined by a single business analyst or shared among a few managers. For example, who are the probable high-net-worth clients to sell them a premium bank account. There can be some consideration like the average account balance etc.

Predictive analytics has been unfolding on a parallel track to business intelligence. With predictive analytics, numerous tools allow analysts to gain insight into “known unknowns”. These tools track trends and make predictions, but are often limited to specialized programs. In the previous example, the probable high-net-worth client could also be the spouse of an existing high-net-worth client that can be figured out using predictive analytics.

Data Science on the other hand is an interdisciplinary field that combines machine learning, statistics, advanced analysis, high-performance computing and visualizations. It is a new form of art that draws out hidden insights and puts data to work in the cognitive era. The tools of data science originated in the scientific community, where researchers used them to test and verify hypotheses that include “unknown unknowns”. Here are some of the examples:

  • Uncover totally unanticipated relationships and changes in markets or other patterns. For example the price of a house based on nearness to high voltage power lines or based on brick exterior.
  • Handle streams of data—in fact, some embedded intelligent services make decisions and carry out those decisions automatically in microseconds. For example analyzing the users click pattern to dynamically propose a product or promotion to attract the customer.

As discussed, Data Science different from from traditional business intelligence and predictive analytics in the following way.

  • It brings in data that is orders of magnitude larger than what previous generations of data warehouses could store, and it even works on streaming data sources.
  • The analytical tools used in data science are also increasingly powerful, using artificial intelligence techniques to identify hidden patterns in data and pull new insights out of it.
  • The visualization tools used in data science leverage modern web technologies to deliver interactive browser-based applications. Not only are these applications visually stunning, they also provide rich context and relevance to their consumers.

Data science enriches the value of data, going beyond what the data says to what it means for your organization—in other words, it turns raw data into intelligence that empowers everyone in your organization to discover new innovations, increase sales, and become more cost-efficient. Data science is not just about the algorithm, but about deriving value.

 

Disclaimer: The postings on this site are my own and don’t necessarily represent IBM’s positions, strategies or opinions.

 

3 Compelling Use cases for Entity Analytics

Entity analytics is used to detect non-obvious relationships, resolve entities, and find threats and vulnerabilities that are hiding in your disparate collections of data. Through the medium of three use cases, let’s try to understand how Entity Analytics can help organizations enhance their customer experience.

entityanalytics1Scenario 1

Entity Analytics can detect non-obvious relationships between entities. It can also analyze new data sources in context leading to new insights and opportunities. In this scenario you have some data in an MDM system and another set of data in a spreadsheet file. Suppose you want to run a marketing campaign to target high-net-worth clients to sell them a premium bank account. The information in the one MDM system in isolation doesn’t give you the needed information. You want to bring these two sources together and determine if you can identify individuals that can be targeted for the new account.

In the MDM system, John Smith lives with Mary Smith. The spreadsheet file shows that John Smyth (spelled differently) is actually a high-net-worth client. Combining this information we can say that John Smith is actually the same person across the data sets. He’s a high-net-worth client, and he has a wife. With this information you want to target Mary Smith with a premium bank account because she lives with a high-net-worth individual. Entity analytics enables you to discover and understand this opportunity.

entityanalytics2

Scenario 2

Entity Analytics can find where threats and vulnerabilities are hiding in big data and respond efficiently. In this scenario for a risk assessor in an insurance firm, severe rainfall is predicted within a geographical area that includes the client’s residential location. When pulling up the client data from MDM and the flood warnings being issued from the environmental agency, we can match across the data sets to identify that a number of properties are at risk. So, the client can then be provided an early warning to help mitigate risk and increase the flood risk value on the client’s property renewal. Also, if you have an elderly customer that is at severe risk; you can take action to notify the emergency service to ensure a proactive resolution to any potential threat.

entityanalytics3

Scenario 3:

Lets see how using Entity Analytics, MoneyGram International Inc.,  a money transfer company gets notified of questionable activities in real time for faster predictive and preventive decision making. This helped them to save $200 million in just two years!

Summary

Entity analytics help organizations by launching more target-oriented campaigns and by reducing the risk of fraud. With the help of entity analytics, organizations can predict and preempt suspicious activity faster and with reduced costs. Entity analytics further help by allowing enterprises to detect the entities that are the same, regardless of whether the entities are hidden or masked. So following questions can be raised:

  • Does this Analytics require an MDM Ninja or can something be set up easily by a Business user?
  • Do we have Entity Analytics available on Cloud for decisions that cannot waaaaiiiiittttt?

Stay tuned for my next blog.

 

Cognitive 1 – Need for Cognitive Systems

watsonThere is a lot of buzz about Cognitive Computing. I plan to write a series of blogs on it as cognition is framing the future of the digital economy. In this blog, lets explore the need of Cognitive technology.

In today’s world  we see the business models converge across categories and industries. The Ubers’ and Airbnbs’ have shifted consumer behavior very rapidly threatening the existence of very strong, established players in the market. This new dynamic was result of their on demand model powered through the immediacy of technology. Now, with emergence of such new business models, organizations can no longer continue to see their competitive set from within industry; they need to be structured to look beyond their traditional boundaries. As brands, businesses and organizations shift to become lifestyle centric, competition can come from anywhere.  Piotr Ruszowski, chief marketing officer, Mondial Assistance, Poland  says,  “The biggest threat is new competitors that aren’t yet classified as competitors.”

To win in this dynamic age, there is a need for organizations to become all knowing. This means getting insights from all data including the ‘dark’ data that sits outside of the firewall of the organization. It includes unstructured information – books, emails, tweets, journals, blogs, images, sound and videos.  The challenge here is that this pool of dark matter is only going to get bigger – the statistic here that “By the year 2020, about 1.7 MB of new information will be created every second, for every human being on the planet” represents the magnitude of how much we really mean when we say big.

As you go outside your firewall to the data that’s coming, it is increasingly unstructured. Traditional systems are programmed and so are not structured to be able to glean insight from dark data. Therefore organizations need to look to cognitive systems that have the capability to be able to make sense of it How? We will explore later, but the key difference here is structured data will tell you that your sales are down for instance, but it’s the unstructured data that can tell you why?

With Cognition, Business technologies that automate and detect can now also advise and enhance human expertise, powering organizations to be able to make richer, more data driven decisions.

Summary:
Here are the benefits of a cognitive business:

  • Puts to work all forms of data, whether structured or unstructured
  • Facilitates evidence-based, confidence-weighted decisions.
  • Discovers new insights and patterns in new kinds of data.
  • Learns and adapts with use, actions, outcomes and new data to stay current.
  • Navigates natural language to allow conversational-style interaction, enhancing adoption and use

The success of cognitive computing will not be measured by Turing tests or a computer’s ability to mimic humans. It will be measured in more practical ways, like return on investment, new market opportunities, diseases cured and lives saved. How cognitive systems work and what are some of the industries already benefiting from cognitive systems? Stay tuned.

 

The Real Opportunity for IoT

In my last blog, I described about IoT and ended with a thought on the “Real Opportunity” of IoT.  Lets explore this a little further, but in one word, it will be Analytics.

Here is a recap on Internet of Things (IoT) : First, in the simplest terms, IoT deals with physical devices that generate data from sensors and send the streams of data via the Internet to some kind of “hub” for data collection, visualization, and analytics. Second, IoT deals with multiple types of sensors and data formats. Third, IoT solutions might deal with thousands and millions of connected devices and huge amount of data.

Now, Billions of Internet-connected ‘things’ will, by definition, generate massive amounts of data of varying complexity, formats and timeliness. This is just a swamp, especially if all you do is collect data and don’t do anything with it.  For example, Insurers pay more than $1 billion in claims in the United States for cars and trucks damaged by hail. Can Weather Company’s weather data make it possible for insurers to send text-message alerts to policy holders, warning them of an imminent hailstorm and advising them of safe locations nearby? Note IoT will make it possible to identify the exact location of these cars /trucks and identify the owner to send the text message!

AnalyticsTherefore while many people focus on the devices themselves— how they function, how they perform and how they look—the real opportunity is in the data these devices are consuming and generating and the value it provides for businesses and even entire connected cities. Retailers will piggyback on Analytics, and use IoT to pull consumers into one of their channels, where they will entice them with products that have been contextualized and personalized for the customers’ gratification. And there will be similar usecases for manufacturers, servicing organizations, public utilities, industrial, telecommunications, healthcare providers and more—to serve their customers in new, personalized ways. Using predictive, prescriptive, cognitive and investigative analytics will make it possible for organizations to discover new relationships and correlations that bring together broader and deeper insights that lead to smarter business decisions in terms of risks, costs, growth, customer service and other things.

What all will be required for organizations to harness the power of Analytics and what will be the challenges? Stay tuned.

Disclaimer: The postings on this site are my own and don’t necessarily represent IBM’s positions, strategies or opinions

Internet of Things (IoT) – Demystified

There has been a lot of talks around IoT. In this blog I wish to demystify IoT a little bit and share some facts and my understanding on this topic.

IoT is not about Internet of connected computers, rather it is about is an Internet of connected devices (or things) that broadcasts loads of data about devices—their interactions with their owners and with each other—that traditionally had little to no computing capacity, but now do.

So lets start with some Facts: 
By 2020, there will be 28 times more sensor-enabled devices in existence than there are people in the world. Of those 212 billion enabled devices, 30 billion will be connected to networks and potentially to each other. These device include everything from cellphones to coffee makers, washing machines, headphones, lamps, wearable devices, and more. A device can also be a component of a machine, such as a jet engine of an airplane or the drill of an oil rig. These smart devices could respond to properties, such as vibration, chemicals, radio frequencies, environment, weather, humidity, light, etc.

IoT_picture_car2So what value proposition will these sensor-enabled devices bring?
Cars with on-board sensors can report back to manufacturers with information on the wear and tear of parts, indicate the cause of system failures and generate warranty notifications.
Store shelves can connect with the supply chain when they’re running low on inventory of a certain product.
Skyscrapers can send building managers information about how much electricity they’re using—and make suggestions for how to reduce it.
Wearable monitors can alert doctors about the side effects of medications and provide patients with advice on how to manage their symptoms at home.
Airplanes can connect with weather stations to help predict turbulence and avoid it during flights.

So what are the challenges and where is the “Real Opportunity” in IoT?
Stay tuned…
[Hint]: 90 percent of all data generated by devices such as smartphones, tablets, connected vehicles and appliances is never analyzed or acted on. Imagine the possibilities if that were increased to 20% or more.