The Data Scientist Who “Listens to the Problem”

My most recent in-flight reading was Thank You for Being Late. In it, Thomas Friedman says risk of AI isn’t that it’s going to take over humanity, HAL-like, but that we as humans could become so entranced by technology that we’ll neglect to teach it human values. It’s not machines v. humans or technology v. creativity. The more technology develops, the greater the opportunity to add to it our kindness, our fairness, and our creativity.

Jorge Castañon, Data Scientist at the IBM Machine Learning Hub and this week’s “You in the Private Cloud,” interviewee, clearly agrees. He and I met this week to discuss math, art, and the future of data science.

What was your first job at IBM?

To understand what data science is. There were so many different definitions! I decided it’s the combination of three things: mathematics, computation, and creativity. You need the creativity to listen to the problem and come up with the math. You need the math because data science requires a very deep understanding of the math that lies behind it. Then you need to compute the solution.

What do you mean, “Listen to the problem?”

Math is like a foreign language that not everyone can speak. When I’m listening to a problem, I’m translating from English to math and then translating back to English to continue the conversation.


The mathematics is distinct from the computation?

Yes. You think of a method mathematically. You eventually need to implement it in the computer: that’s the algorithm part. But first, it’s you and a blank piece of paper, and your thoughts, and eventually a math solution. The first person who thought about linear regression or least squares, that person was mathematical. It was a bunch of data points in space, and then, “Let’s find a model that fits those points” — but first it was math.

IBM was named on Gartner’s Magic Quadrant for Data Science Platforms for 2017, because of DSX with machine learning, and also the work you and the team are doing. A lot of it is side-by-side with clients: what’s that like?

It’s super fun! Learning about new problems is the best part of data science. The minute you start a conversation with a domain expert, to see what are the important parts of the model, what you can use for your math solution: that’s the exciting part. Talking to customers is a way to find the most interesting problems to solve.


I would imagine coming out of Rice with a PhD in Computational Mathematics that you had a lot of career choices. Why did you choose tech and why IBM?

Rice University is in Houston, so there were opportunities in the financial world and the energy sector and a lot of money to be made. I went to a conference and met IBM recruiters and got good indication of the spectrum of expertise at IBM. I felt I would be able to go wherever I wanted to in terms of the research and technical challenges; I would not be limited to one narrow role.

What’s the one thing about work that you are most excited about?

Collaboration. As a computational mathematician, you know a lot about certain things. But to go and talk about energy efficiency, or credit unions, or TV marketing, that gives me new topics where I can apply math and make a difference: to health care for example, or by making a building more efficient.

You are working at the edge of technology that doesn’t quite exist yet.

Definitely. My first project was to identify what is data science: that was unstructured. Then, how to use data science in our products: unstructured. How to apply machine learning: unstructured. It’s very exciting, to find the structure of things that are amorphous or not yet reified. And that’s what mathematics is. It goes back to my whole path, to the creative problem-solving that drives me.


Where do you see data science going? Is it part of the machine learning path, or will it diverge?

It’s an open question as to whether data science is going to be automated and humans won’t be needed. I think they will be.  The creativity aspect of data science cannot be automatic.

What do you do for fun outside of work?

I love art, and traveling with my wife: she’s also an applied mathematician. We got to museums and I take photographs of art. I used to do life drawing, but after the PhD and work — you get busy! I love James Turrell in particular; his work is based on what he called “the geometry of light” and he studied math in college.

Customers tell me it’s not just our skills they appreciate, it’s the commitment we make to their success, and they see that from working directly with you and IBMers like you. Thank you.

You are welcome. It is a pleasure to work here. I have a lot of space to grow.

Name: Jorge Castañon

Years at IBM: 3

Home town:  Mexico City

Currently working on: IBM Machine Learning Hub

All-time top five artists:

  1. Francisco Toledo
  2. James Turrell
  3. Willem De Kooning
  4. Mark Rothko
  5. M.C. Escher


Dinesh Nirmal,

Vice President Analytics Development

Follow me on twitter @DineshNirmalIBM

IBM Machine Learning for z/OS – Like no other

Like no other Private Cloud

With many of the top banks, retailers, and insurance organizations using IBM® z Systems® , combined with tried and tested virtualization capabilities, EAL5+ security rating and the ability to handle billions of transactions a day[1], the platform becomes attractive as a private cloud for running advanced analytics as well as cloud managed services.

Those organizations are in an enviable position, with volumes of new and historical business-critical data available on such powerful and reliable systems. The sheer volume and velocity of the transactions, the richness of data in each transaction, coupled with data in log files, is a potential gold mine for machine learning applications to exploit cognitive capabilities and do smarter work, more intelligently — and more securely.

Leveraging Machine Learning on z Systems

Set against an asymptotic curve of information growth, Chief Information Officers and data scientists constantly battle to gain deeper insights from the volumes of transactions and log data on the platform (and many other platforms) and turn those insights into concrete gains. In most cases, the CIOs already have astute teams of data scientists and data engineers combing through this data — and yet they see their teams struggle to make enough time for the deep work they’re trained to do.

“Enterprises are well aware of the tremendous potential and value of the transactional and operational data on their z Systems. Yet most of them struggle with how to expose the data within the enterprise in a secure and controlled way thats still flexible enough to foster innovation and support efficient access for a variety of roles data scientists, operations, and application developers. Not an easy task, but organizations that can do so potentially obtain an edge over the competition.”

—Andrei Lurie, DB2 for z/OS Architect, IBM

Machine learning has the potential to be the perfect intelligent app — to hike efficiency, create and cement deep personal relationships with customers, push into new lines of business and new markets while helping to minimize financial risk and fraud.

I have heard customers say that the mainframe has never been hacked. But it doesn’t mean cyber criminals aren’t trying, nor that unscrupulous people aren’t attempting to commit fraud. Having applications that embed predictive models that can analyze, sense, react and become smarter with every transaction and interaction in such a business critical environment brings us a long way toward identifying and preventing potential fraud.

But z Systems is not just about transactions. It is already considered to be a hybrid transaction and analytics processing (HTAP) environment with a complete set of the analytics capabilities and acceleration technologies available today. IBM has also added full exploitation of Apache Spark™ on both z/OS and Linux® on z Systems – a solid base for building, testing and deploying high performance machine learning applications.

“By running advanced Apache Sparkanalytics directly on their production systems, enterprises can improve both the efficiency and timeliness of their insights. Moving Spark inside the mainframe also simplifies and can help reduce security risks as there is only one copy of the data to protect, and that copy resides inside z/OS’s security rich environment.”

— Fred Reiss, Chief Architect, IBM Spark Technology Center

For all these reasons and more, we are delivering the full range of our machine learning capabilities to z/OS essentially bringing advanced ML to the world’s most valued data.

Machine Learning without Compromise.

When asked to describe machine learning I break it down into three perspectives: Freedom, Productivity and Trust. I find these resonate well with customers’ needs.

Freedom. Think of freedom as a set of unified but powerful capabilities such as the flexibility of the interfaces that can be used to interact with machine learning — whether a Jupyter notebook or intuitive graphical interfaces catering to the needs of various personas from beginners to expert data scientists. With support for Python™, Java™, and Scala, different organizations can leverage their preferred programming language and skills when building machine learning applications. Machine learning from IBM can be developed on and deployed across different computing environments such as private cloud and public cloud – including IBM z Systems z/OS with a choice of frameworks such as SparkML, TensorFlow™ and H20.

With the data available to machine learning solutions, users can create advanced algorithms or choose from a set of predefined powerful algorithms and models without requiring advanced data science expertise.

Think of all this capability running on one of the highest performing platforms available: IBM z Systems. It means machine learning can be brought to bear many thousands of times per second[2] — which can help reduce costs and risks, finding and leveraging new opportunities at every transaction and interaction.

Productivity. To make machine learning consumable it has to be easy and intuitive for end users. To this end, IBM machine learning was built around three core principles of simplicity, collaboration (across multiple personas) and convergence of a wide range of technologies from our analytics portfolios and our research laboratories. The user experience is key, whether the user is a data scientist – advanced or beginner — or a computing generalist. Across personas, IBM Machine Learning lets users engage and collaborate on machine learning projects– leveraging the combined skills of the team. Wizards within the tools provide step-by-step processes and workflows that automate many aspects of building, testing and deploying learning machines. As part of the process the IBM Cognitive Assistance for Data Scientists (CADS) automates the selection of the best algorithm given a training data set. It starts by allocating a small part of the data set to each candidate algorithm, then estimates performance on a full data set. It uses the estimate to rank algorithms, and allocates more data to the best ones. It iterates until the best algorithms get all of the data set.

Trust. Once a model is built and tested, it needs to be deployed. A model – in fact the entire machine learning application (learning machine) — is similar to a living organism, evolving and adapting over time as it encounters new data with each transaction and interaction. This is achieved through a continuous feedback loop that enables the model to adapt and change, altering specific parameters within the model itself to become smarter and more consistent over time – while avoiding overfitting.  This auto-tuning is key to reducing manual intervention. Of course some human intervention or model adaptation may be necessary where a human judgement or decision is required. Therefore, keeping track of the version of the models over the lifecycle of the learning machine is important for audit purposes or to fall back to a previous version.  Another aspect of trust is of course the governance and security (the who, how, when, where) of the data, the models, and the many machine learning artifacts. IBM z Systems is recognized in the industry when it comes to security[3]– and a key reason why some of the biggest names and well known organizations across many industries run their business critical applications and data on the platform.

These three perspectives are summarized in the Figure #1 below.


Figure #1 IBM Machine Learning the complete picture.

From a technology point of view, our aim is to free up data science teams to do the deep work that’s being asked of them — work that gets harder and harder as the world moves faster and with less certainty. Ultimately, the gains that CIOs are seeking will come from a collaboration between smart data systems and smart data scientists. Machine learning on z/OS will help enable and encourage that collaboration.

IBM Machine Learning Hub – Beyond the Technology

While the technology aspects may deliver very advanced machine learning capabilities, IBM recognizes the need to nurture and partner with organizations as they embrace and fully exploit its machine learning technologies. The first IBM Machine Learning Hub will provide the means to achieve this, with the aim to accelerate and enrich organizations’ machine earning knowledge and expertise.

The “hub” will allow organizations to access IBM world-class data science professionals who can provide education and training, expert advice on all aspects of machine learning – as well as lead and deliver proofs-of-concept and full client engagements. They focus on delivering tailored machine learning knowledge and skills transfer built around the needs and wants of customers. This combination of both the technology aspects and the knowledge / skills base is an opportunity to provide a unique machine learning experience what I consider to be the machine learning ecosystem.

Let me close this blog post by inviting you to take a look at a short video on machine learning here and reading the recent announcement of IBM Brings Machine Learning to the Private Cloud .


Dinesh Nirmal,

VP Analytics Development

Follow me on twitter @DineshNirmalIBM



[2] based on IBM SPSS Modeler Scoring Adapter for DB2 for z/OS performance

[3] EAL5 + Security Rating

Apache Spark, Spark and the Spark logo are trademarks of the Apache Software Foundation (ASF).

Linux is a registered trademark of Linus Torvalds in the United States, other countries, or both.

Python is a registered trademark of the PSF.

Java and all Java-based trademarks and logos are trademarks or registered trademarks of Oracle and/or its affiliates.

TENSORFLOW is a trademark of Google Inc.

IBM, the IBM logo, and are trademarks or registered trademarks of International Business Machines Corporation in the United States, other countries, or both.

Database and Machine Learning expert, Lead Vocalist on “Dance with Me”

Ketki Purandare: You in the Private Cloud: A bi-weekly series of conversations with IBM talent around the world

The people in IBM Private Cloud come from so many different countries, backgrounds, and traditions that we can sometimes feel like abstractions to each other. These interviews are a chance to correct that feeling — a chance to focus on our common goals and endeavors, and enjoy our diversity.

On a recent trip to Boston, I convinced Ketki Puranare’s manager to free her up from user interface (UI) development long enough to talk with me.

You’ve been a full IBM employee less than a year, but you actually started at IBM in May of 2015 as an intern on IBM InfoSphere information Server (IIS), working in the data governance team. What was that like? I was one of five interns working on a hybrid on-prem solution for data standardization. IBM has a good vision in this area: We could all learn something new, and the product could benefit from the latest technology. We chose AngularJS as the framework for the front end, and for most of the internship, I worked with the Data Standardization team to develop a service to view and add standardization rules.


How did it go? Great! I was able to create a working proof of concept (POC) and did some fun stuff like customizing Apache Tika™ with a csv parser so users could upload a csv with standardization rules. In the second half of the internship, I worked with the Information Analyzer Thin Client which involved writing a REST layer to use existing functionality of the on-prem solution and make it available on a browser as well as developing the entire UI in AngularJS. I learned a lot under the mentorship of our architect— especially since I got to be present from the inception of the project and attend the Design Thinking sessions.

It sounds as though you experienced a true IBM immersion. I realized right away that this is a place where I have a lot of opportunities to grow by working with such experienced people and trying out new challenges with different technologies and tools. There’s so much pooled expertise, it’s inspiring to be a part of it.

Tell me how you decided to study computer science. When did you first know it could be a passion? I wasn’t totally into computers as a child and actually aspired to be a medical student. Then my older brother chose Computer Engineering as his field of study, and I started to get interested. I must’ve been about 14. Sometimes he’d ask me to try out a snippet of code in C/C++ using logic. When I was able to solve it I felt awesome and I think that was what got me interested in pursuing computer science myself.

And it sounds like you’ve been working with logic and algorithms ever since. I know you got your master’s degree in Advanced Databases and Machine Learning. How did you like the program at SUNY Buffalo? The master’s at SUNY Buffalo was a thrilling experience! We had some of the best professors and worked in small groups or even individually on projects like building an indexer and retrieval system similar to Apache Solr™.


Where do you see yourself 10 or 15 years from now? My ambition for the long term is to be a software architect or a Fellow/Distinguished Engineer. I’ve done this at a small scale in my master’s’ projects, but I want to provide blueprints for larger solutions. We have a speed-mentoring initiative here at the Littleton lab which I haven’t taken advantage of yet, but that’s at the top of my list right now.

Would you say you have a philosophy around coding and work? A professor at SUNY Buffalo used to say you never get a project right the first time. Most large projects need to be scrapped and redone to improve the architecture. I had a similar feeling after completing projects in grad school where I’d think I could’ve done it better had I done it another way. You can only get that sense from experience.

I know you spent time in Pune, in the west of India, not far from Mumbai. It’s one of the fastest growing cities in Asia but also really known for research and education. What’s it like? Pune is known as the Oxford of the East! It’s a great city and attracts a large number of students from all over India every year for its quality education. 

Speaking of culture, what do you do outside of work? I used to be the lead vocalist in a band in my school days back in Pune. We used to have concerts and even compose music for local movie productions.


What kind of music did your band play? Fusion music like Indian Classical with Rock. There’s a song called “Dance with Me” that I sang in Marathi as part of a movie called “Mission Possible” in 2010. We were still undergrads back then.


Thank you, Ketki. You’re obviously destined for great things at IBM and I feel so lucky to have you with us on Private Cloud.



Name: Ketki Purandare

Years at IBM: 1

Home town: Pune, India

Currently working on: UI for Information Analyzer

All-time top five songs:

  1. Bring Me to Life — Evanescence
  2. Sweet Child O’ Mine — Guns N’ Roses
  3. Chammak Challo — Akon
  4. New Day has Come — Celine Dion
  5. Vande Mataram — A. R. Rahman


Dinesh Nirmal

Vice President Analytics Development

Follow me on twitter @DineshNirmalIBM


Apache, Apache Soir, Apache Tika, Apache are trademark of the Apache Foundation

Welcome to the Private Cloud

Readers of this blog know that I like to imagine the world through the eyes of my young son. I’m struck by his constant drive to push himself to his next edge of independence. I also know his appetite for danger goes only so far. He understands some of the safety boundaries we have set to protect him from the chaos and to help him thrive. Home is a safe environment in which we prepare him for the outside world, public places like school where he interacts with other students, sharing resources with others.

His effort to find the right balance of exploration and safety resonates with what we mean by “private cloud” and preparing clients for a hybrid cloud environment: private plus public, as in figure #1 below:


Figure #1: Hybrid cloud the path toward optimal business outcomes

Hybrid cloud can provide ultimate flexibility by allowing organizations to place data and associated workloads where it best makes sense – for optimal business outcomes which I discuss in more detail later in the blog.

Private Cloud defined

In the simplest terms, private cloud (sometimes also called internal cloud, dedicated cloud or corporate cloud) provide all the benefits of cloud provisioning, management capabilities along side the scalability, agility, and the developer-driven services available from cloud vendors — but behind the firewall. Figure #2 below offers some details about the differences.

Public and private clouds are both destinations for the execution of business workloads. More and more, we see organizations eager to take a hybrid approach which allows those workloads to seamlessly execute “together” across public and private cloud allowing those customers ultimate flexibility based on (but not limited to) :

  • The volumes and types of data
  • Sensitivity of the data
  • Performance and service levels required
  • Security requirements
  • Business criticality
  • Data regulation and governance
  • Types of systems, processes, and applications

How you put the pieces together depends on the needs of your business. There are many economic and service level factors to consider. A private cloud is often the responsibility of the organization running it. Besides the factors mentioned above, the responsibilities can include: hardware, software, support, maintenance, service-level agreements with the business and all the necessary human and technical resources associated with it. With a public cloud many of these economic and service level responsibilities can often be devolved to a third party – allowing the organization using the public cloud to focus on its core business processes and needs.

That said, some enterprise customers are seeing that many of the benefits typically associated with the public cloud — lower cost, speed of provisioning, reduced management — are increasingly available on private cloud configurations that also allow data to be governed securely, smoothly, and transparently.


Figure #2 : Private and Public cloud differentiation

Life behind the firewall

What we mean by “behind the firewall” depends on individual clients and their needs. It might mean that the data is maintained completely within a client’s own protected data center by the client themselves. Or, that the data and apps live on fully dedicated bare-metal servers off-site, supported by a cloud vendor like IBM managing hardware, maintenance, connectivity, redundancy, and security on the client’s behalf, all of which help that client drastically reduce capital expenses for the servers, in-house IT staff and the burdens of obtaining and updating software.

Avoiding expenses and hassle is just the beginning of what’s possible, but let’s first consider why maintaining a private cloud while exploring public cloud options is the right fit for so many of them. Broadly, private cloud configurations can address two particular needs:

  1. The need to create a highly secure and reliable home for sensitive data, to perform advanced analytics, and to maintain data sovereignty — while allowing that data to be in conversation with data and analytics that are accumulating in the public cloud. In this sense, private cloud is one end of a private/public cloud hybrid configuration in which data is accessed, moved, and managed using secure, service-layer APIs.
  1. The need to modernize systems and processes — even behind the firewall. Organizations who see the benefits of maintaining a private cloud nevertheless demand the clear advantages of public cloud I mentioned before: elastic scalability, agility, consumability of API-driven services, easier management, and rapid provisioning, to name just a few. The key concepts here are:
  • virtualization — The use of virtual operating systems and highly elastic virtual processing power.
  • federation — The ability to take several different physical entities and represent them as a single logical entity.
  • data fabric — A software-defined approach for connecting disparate storage and management resources across private and public cloud. The approach enables multiple components to interoperate through a set of common, standardized services and APIs regardless of physical location, type of data, or type of service. As mentioned above, clear data governance is particularly crucial in hybrid environments — and even more so when country-specific compliance rules require different data policies across geographies.

As my colleague wrote:

Private Cloud is about delivering an elastic data fabric behind the clients firewall. From a user perspective, the experience goes from “Provision me a database to do xyz” to “Here is my data and analytical needs, please help.” There is no need for dedicated repositories for a specific application and user needs are met automatically, with limited human intervention.


Figure #3 : Hybrid cloud architecture  

Path to Cloud Benefits

Regardless of their focus, organizations are hungry for simplicity, transparency, and the ability to move toward cloud without starting from scratch. They know that their future success lives at the edges of wide networks, at the points of direct contact with customers and the outside world. Mobile phones, IoT sensors, and other connected devices are the new lifelines to current and potential customers, who passively or actively exchange volumes of data with servers. That data runs the gamut in terms of privacy and sensitivity: from the temperature of the toaster to credit card information, from glucose levels to the current whereabouts of my son’s backpack. All that activity at the outer edges of the network has shifted a portion of the business into the cloud even for traditionally cloud-wary sectors like finance, government, and healthcare. For those organizations, a private cloud offers an environment for core-mission, transactional workloads even as the public cloud allows them to explore CPU-intensive or streaming applications that are (for now) less central to the business. Not surprisingly, these sectors are exploring tunable hybrid cloud infrastructures. Figure #3 above offers some perspective.

Alongside the need to stay connected to customers, pressure to come to the cloud is also intense in terms of cost savings, easier management / provisioning, and — perhaps ironically — security. Security threats evolve so rapidly and attacks come from so many directions that internal security teams can struggle to keep up. And since some of the most severe cyber-attacks can come from within a company’s own ranks rather than from exterior bots or hackers, the internal teams are finding that the security of the cloud providers can be advantageous in terms of speed, currency, and completeness. As Cameron McKenzie points out, “Enterprises are starting to seriously consider the cloud as a viable option because they’ve realized that security is a battle they can’t win on their own.”

Advantages of IBM Private Cloud

Right now, IBM Private Cloud can help provide the best of the public and private cloud worlds. In fact, a recent InformationWeek post about private cloud states that “IBM is the market leader.” Our deep, in-house knowledge can help organizations breathe easy in terms of performance, cost, security, and white-glove attention and support. We start with the assumption that those organizations need to leverage the systems and processes they have in place by cloud-enabling their investments — rather than starting from square one.

Think of the IBM Private Cloud as a stack. You still need that physical infrastructure that offers high availability, scalability, performance — a strong data and analytics foundation to ingest, prepare, wrangle, discover and transform data into trusted assets. On top of that you need the ability to manage existing investments in applications and solutions as well as creating new services and apps that are cloud-enabled and can be rapidly provisioned – everything from management of the infrastructure to a collaborative development environment. Oh, and the need for security and governance of the data, transactions and applications over their lifecycles doesn’t go away.  All these layers in the stack (regardless of whether an organization buys into all of them) can be provided by IBM today – and many of them were well established and available before the mainstream adoption of cloud.

Customer environments without exception are multi-vendor, consisting of an array of heterogeneous platforms. That’s why the private cloud platform is designed to co-exist and integrate with many different technology infrastructures. The goal is to bring cognitive analytics capabilities to wherever the data is with flexibility in mind – such as delivering offerings in multiple form factors to help meet the diverse needs of our clients on their cloud and cognitive journeys. A great example is the use of Docker images that make it possible to run our analytics and other offerings across many different infrastructures leveraging the attributes of private cloud.

Innovation and Investment for client success

We’re innovating and investing on clients’ behalf to help bring them not only the expected benefits of the private and public cloud, but with the robust internal partnerships with IBM Power and IBM z Systems, business partners like the ones described above, and access to market-leading data management solutions, world class descriptive, predictive and prescriptive analytics solutions – all in a cloud-enabled integrated, secure and governed environment. All this comes together within the private cloud data platform with tried and tested infrastructure, governance, security, data fabric capabilities and cognitive computing services – with the flexibility to provision data and policies across private and public cloud environments. This is an optimal hybrid model.


In subsequent posts, we’ll look at private cloud strategies related to data repositories, analytics, content management, and integration/governance issues — and how these strategies braid together.

In the meantime, I encourage you to click the IBM private cloud page – a great place to explore and try some of the capabilities that exist today, and get a preview of what’s coming soon.


Dinesh Nirmal, Vice President, Analytics Development.

Follow me on Twitter @DineshNirmalIBM

Morgan Fritz: Industrial Design Grad, Fashion Designer, DSX Veteran

You in the Private Cloud: A bi-weekly series of conversations with IBM talent around the world


Our Chief Designer, David Townsend, takes new designers aside when they join our team and tells them, “This is the most complicated thing you’ll ever design. Don’t be afraid to ask dumb questions. And know that everything else in your career you design after this will be easy compared to what you’re about to do.” This week, I talked to Morgan Fritz, a new designer in our San Francisco Studio, to find out what it’s like to turn some of most abstract and complex software in the world into a product that should be easy, and a pleasure, to use.

Why did you choose to study Design? I studied Fine Arts in high school, but it was too much time alone in a room. I wanted something more social, and something for a bigger cause. I went to Carnegie Mellon to study industrial design and then user experience (UX) – I loved the system thinking. You’re thinking, not just making pretty pictures. IBM is perfect for that.

Yes, ‘Think.’ Exactly! The only downside is that is can be hard to explain to other people what we do, because it’s technical challenging.


You’ve been here at IBM for 7 months, and your first project was the Data Science Experience (DSX)v, which went from a seed project to full blown GA in a very short period of time, and every customer I’ve shown DSX to, loves it. It’s a tremendous achievement. What was it like to design it? 

Crazy. We came in, and for all of us it was either our first job out of Design School or we were new to IBM, and it was like, ‘Here’s what we’re doing, here’s what a data scientist is, go!’ The pace was what we wanted: fast, like the momentum at college. And to experience a full product from research to launch for my first project on my first job, to be able to see something go all the way through and to be able to point to it and say, “I did that,” that was a crazy opportunity.

Design plays such a critical role in our products. Earlier today I talked to our machine learning team about how we can make sure design is part of everything we do, to help ensure ease of use. From your experience, do you think design is incorporated in a way that will result in a product that’s easy to use? Oh absolutely, but a lot of that comes with growing and learning. The DSX design process started to work after we learned to work with our dev team and with offering management. After GA, we did a retrospective – how could the process have been better? It happened so fast – files were flying around – so we learned so much from the retrospective.

What do you do outside of work? Travel, most recently around California; I just got back from Yosemite. Next is Pittsburgh. I’m launching a fashion line and I have to show a dress at the Carnegie Mellon Fashion Show. This weekend I have to buy the fabric and make it.

What is it about travel you like? The novelty of new landscapes, or more the process of getting there? Insights. I spent 6 months in New Zealand and I carried a sketchbook everywhere. It was nice to be able to spend that much time there because I made friends and then stayed with them in different places. And even in a work context, we design for a certain market, but we have international studios; we’re designing for a larger base than just people from Silicon Valley, CA, so it’s nice to actually understand different places and people.


What gives you the utmost satisfaction, the product being released, or looking at the design and ensuring that it’s simple enough to use? If you look at satisfaction from an engineering perspective, you’ve coded it, you’ve designed the architecture, and when you release it, that’s the point of satisfaction. There’s no one right answer in design–but there’s this magical moment when you’re like, I think we got it right, I think we nailed it. It’s that point before launch, when we hand off to devs, when we’re not just saying, ‘here’s our design,’ but talking through technical aspects with them: you feel like everything falls into synch.

You’re a group of new talent at IBM. You’re in San Francisco, and there’s a lot of opportunity. What’s is the perception of working here as opposed to, say, at Uber or Lyft? On one side, it’s definitely challenging. Design is just emerging into IBM; we’re still finding a good rhythm to working with developers. On the other side we’re given so much influence. I think that’s so substantial compared to other tech companies, the fact that we can say our small team is part of a core part of this huge product we released. It’s crazy to feel like you have that much influence in such a large company. Looking back on DSX we can say, ‘We did this, everything we designed had an intention, and that intention got followed through to the end.’ That’s such a great feeling.


Name: Morgan Fritz

Years at IBM: <1

Currently working on:  UX for DSX and IBM Data Platform

Hometown: Santa Rosa CA

Top 5 Design Destinations: Queenstown, New Zealand; Santa Fe, NM; Istanbul, Turkey; Santorini, Greece; Santa Rosa, CA

Dinesh Nirmal,  

Vice President, Analytics Development

Follow me on Twitter @DineshNirmalIBM

Phu Truong: Humble Leader, Loves Logic, Hates Calculations

You in the Private Cloud: A bi-weekly series of conversations with IBM talent around the world

If you’ve seen the movie “Hidden Figures” — and if you haven’t I highly recommend you do, and not just because IBM is a central character — you’ve seen how the race to get a man into space was profoundly affected at the 11th hour by one courageous woman, and the help her boss, her friends, her teachers, and her family gave her to get to that one minute in time when she made a difference.

People first. To build rockets to the stars and machines that think, people need to dream things up, and work with sustained, supported effort to make them real.

We have many talented people in IBM Private Cloud. This year, I’ll continue to meet and talk to as many of you as I can and I’ll post our conversations here every two weeks. My hope is we’ll get to know each other, and feel even more connected and supported in our work.

This week, I was able to lure Phu Truong away from coding on the IBM data platform to meet at IBM Silicon Valley Laboratories, San Jose, CA.

You were an intern with us until just a few months ago. Why did you choose to come to IBM full time, out of all the choices in Silicon Valley? The appeal of IBM is the opportunity to work on new technologies, specifically, new technologies on the back end. A few weeks into my internship the senior engineer I worked with set me to work on learning Node.js® and React. I want to be a full stack engineer so now I’m working on UI, but to be really great there you need a feel for art, and I don’t have that. The back end is pure logic. I loved it, so much so that I started staying very late at night to work.


Some people love their jobs because of people, or culture, but clearly, you love the technical work. How did you decide on computer programming as a profession? I come from Vietnam, and I had no programming background there, to be honest. I studied mathematics at university and planned to go into it professionally, but I’m very bad at calculations. I make mistakes all the time! What I love about mathematics is logic — the feeling I get when I solve a problem using logical thinking is intensely satisfying to me. I feel very good about myself.  So when I came to the U.S., I had a fresh start. I asked my friends to help me find a field that uses logical thinking to solve problems, and they recommended computer science. One week into my first CS class, Data Structures and Algorithms, I knew I’d found my profession.

So now you’re at IBM, you worked on the Data Science Experience (DSX) and now you’re working on the IBM data platform. Are you thinking of following the full path from engineer to Senior Technical Staff Member (STSM) to Distinguished Engineer (DE) to IBM Fellow? I don’t know, that may be too much!


I hear great things about you so maybe not! You’re already mentoring others in your team on Node.js, after being here only a few months. I consider it more like sharing knowledge. When a colleague comes to me with a question, I might know something they don’t and they might know something I don’t. I might say something wrong when we’re working together and that’s an opportunity for them to correct me and for me to learn. Growing up, I helped my younger brother with his schoolwork, so I guess it’s natural for me to help. But it benefits everyone.

What do you like to do outside of work? I like to play Ping Pong with my friends from San Jose State, or go with them to the beach. And I love to travel—I want to go to Cancun, because of all the natural landscapes the beach is my favorite and I’ve heard it’s spectacular there. After that, Paris and London. I love eating out, so much so that I tell my friends I want to marry a chef!

You have an adventurous spirit! IBM is an international company so, I don’t know about Cancun, but travel to Europe is likely. What’s it like living in the heart of Silicon Valley after growing up in Vietnam? I grew up in Saigon, in a very tall, very thin town house: Saigon is famous for thin houses. Here, being surrounded by rolling green hills and close to the beach is wonderful. I think my family worried about me when I moved here, not that it was dangerous, but that I might just chase money and give up on my education: I worked as a waiter, a data entry clerk and a school bus driver, any job I could get, I took. But I never gave up on my education. I think now they don’t worry about me anymore. I think they might be proud of me.


You’ve achieved a great deal here in a very short period of time, making a significant contribution to two products that customers like. It’s tremendous, and I’m happy you’re here. I am as well. I think the biggest difference between Vietnam and here is in education and learning. In Vietnam, education was driven by memorizing things and was not interesting to me. And, we are taught to do exactly what teachers tell us to do; they don’t give students a chance to explore their interests. So to be first at San Jose State and now at IBM where it’s part of my job to learn new skills—well, I like it very much.

Name: Phu Truong
Hometown: Saigon, Vietnam
Currently working on: IBM Data Platform
Favorite Programming Language: Node.js
Top 3 travel destinations: Cancun, Paris, London
Best Vietnamese Food in Silicon Valley: Pho Y 1 on the Capitol Expressway, San Jose

Dinesh Nirmal,  

Vice President, Analytics Development
Follow me on Twitter @DineshNirmalIBM

Node.js is a trademark of Joyent, Inc. and is used with its permission. We are not endorsed by or affiliated with Joyent.

Talking with IBM Talent around the World

Introducing Kewei Wei, from Beijing China

My founding principle for our organization is, “people first.” I believe that to build great products, you start with talented people and invest in their work and their individual well-being. The great products they make will in turn bring customers.

One of the best things we get from work is the satisfaction of being part of a group—bringing a divers set of skills together to accomplish something that no individual can do alone. We’re wired to feel good when we act together towards a common goal—Sebastian Junger (author of “The Perfect Storm”) writes in “Tribe” that interdependence and community are required for human happiness. If you played in your school orchestra or soccer team, you know the feeling; I get it with my running buddies when we silently sprint the last half mile, pacing each other to the end of the trail.

In the IBM Private Cloud team we have almost 3,000 people, and a list of clients who represent  a significant piece of the financial, government, and commercial world. And it just so happens that we sit at a “tipping point” with customers wanting the advantages of all that cloud offers, plus premiere machine learning technology, in a private cloud environment.

As a team of thousands working in labs from Austin to Zurich we have grown to know each other beyond the sometimes less than energizing interface of email.

I decided to conduct a series of interviews, You in the Private Cloud: look for them weekly, here. It’s a great way to bring new information forward that increases our collective knowledge.

People who are talented and engaged tend to have interesting lives outside of work—hobbies that test limits and relationships that matter, so I’ll be asking about that as well.

To kick off the series I talked to Kewei Wei last week in Beijing. Kewei is the lead developer for Machine Learning on z Systems in our IBM China Development Lab.


Kewei and I had just come from a meeting with a customer – one of the largest financial institutions in the world. They had just told us they want to be part of the closed beta of our machine learning offering on  z Systems – a testament to the tremendous work Kewei and the team have done. We sat down in a quiet part of the office and I asked him what he thought:

Kewei: This is exciting news! We weren’t sure how quickly this customer would move into new technologies.  But I believe they were, to a certain extent, feigning reluctance; they’ve been trying all along to find a balance between moving ahead with machine learning and other new technologies, and keeping the stability and security that they absolutely need to maintain. They cannot take risks with financial information, given their global position and their scale. Now we have an opportunity: if we can show them during the closed beta that we have the ability to be both of these things to them—a protector of security and stability, as we always have been, and now also the conduit to state of the art machine learning technology, they’ll take the risk and move forward with us.

You worked on z Systems for 10 years. Then, 3 months ago, out of the blue you’re told you’re leading Machine Learning on z Systems , and you have 90 days to deliver. How was that work day for you?
To be completely honest, my first thought was, “This is impossible”. In ten years in the z world, we’d never done anything like this; our release schedule was more like 3 years. But we did it, and I believe it’s because we were all passionate about machine learning. We know it’s the technology that’s going to make the greatest change for the world, and that’s something we all wanted to be part of. So we put our heart into it. We learned the open source libraries, the skills we needed to deliver the new micro-services structure, and, critically, we learned how to prioritize. With so much to learn, if we hadn’t done that, we wouldn’t have been able to deliver.

Why is machine learning so exciting to you? What would you like to solve with AI?
I’d like to build machine learning AI to help us perform systems facediagnoses and fix defects. That way, we humans could have time to focus on how to make our world better, instead of just making things work correctly.


A fine distinction. Tell me your take on the Private Cloud.
I think it’s a good thing. We’ve been hearing about the importance of public cloud for a long time. Customers  are telling us they will not be ready for the move to public cloud in a short period of time. But they’re also telling us they can’t afford to wait years for security to mature in the cloud. They need machine learning now, behind the firewall, to be competitive.

Private Cloud represents the reality of the market—what our clients are asking for. It means we’re helping them move forward, but without having to face some of the perceived risks of some public clouds.

I also believe innovation in machine learning will happen in Private Cloud. The foundation of machine learning is, of course, data, and as of today, the majority of critical data is still in the hands of customers who keep it behind the firewall. And customers know they have to use machine learning in their core business, which is behind the firewall. They’ll invest in private AI, and we’ll get to build that.

In terms of innovation, what do you see in China that might come to the US from China’s hyper-evolved app technology, or from the broad consumer adoption of virtual reality there?
Honestly, looking at from technology perspective, there is nothing new in what China is doing today. I think the key is to move fast. Don’t wait, listen to consumers, be willing to take risks, dare to fail and always move faster than your competitors

Where did you grow up, and how did you come to be an engineer in Beijing?
I’m from Jinzhou—it’s a city of about 3 million people (small, in Chinese terms) in Liaoning Province in the northeast of China. My father was a bus driver and my mother worked for an oil company. I am very proud of them! When I was 18 I came to Beijing and enrolled in University. In one of my very first classes, a professor helped me write a Pascal program to simulate a game of chess. I was shocked that this was possible—and in a way I never stopped being shocked, excited, I mean, by what we can do with programming.

What do you like to do when you’re not programming?
Read! I love history books. I just finished “Zhang Juzheng”, I a biography of a famous Chengxiang –a prime minister- in the Ming Dynasty. I’m looking for a book on European history for my next one. I bought Gibbon’s “The History of the Decline and Fall of the Roman Empire” but a few chapters in I realized I’d better start looking for a thinner version of it.

Realistic. I also love to travel with my wife and my 5-year-old son.  I think the most fun thing in life is being in a beautiful place, in the mountains or in a relaxing beach city like Xiamen, with the people you love most.


What do people talk about in China when they talk technology? Are people concerned about security and privacy? Or Artificial intelligence displacing the traditional work of people in the labor market?
What concerns individual consumers far more than security in China is convenience and lower cost. For example, the latest hot business here is food delivery, people love it that we can order a food from a nice restaurant from an app and get it within 30 minutes, for less than we’d pay to eat in the restaurant. Part of the reason it’s so low cost right now is because China’s internet companies are investing without regard to cost: they want to lock in users for long-term return.

It’s almost the inverse of the concerns of government or large enterprise: individuals choose convenience over security every time. For example, you can pay by Alipay or Wechat almost everywhere, including places like vegetable markets where POS probably will never exist. Alipay and Wechat are not as secure as credit cards, but they are far simpler to use. It helps that Alipay and Wechat have committed to compensate consumers if you lose money because of any security defects.

As for AI displacing traditional labor, people in China don’t worry about it much. AI is still a pretty new concept to Chinese. What we know is just AlphaGo or Waston. It still sounds like something far off in the future, not close to our daily lives at all. But AI is starting to draw more and more attention, so this could change soon.

Overall, tech in China is booming, and companies like Alibaba and Huawei are knocking at your door. Why stay at IBM, with those kind of opportunities?
Oh, that’s an easy one: the people. In my opinions you can’t find a better company in the world for talent. I have the greatest people around me, and it’s because of them I had the confidence that I could deliver this project on time.

Kewei, having you on the team has been a blessing: you have been a true IBMer, and delivered machine learning on z Systems in such a short time , an impossible task. Companies would probably not buy our products without people like you. Thank you.

You are welcome. I enjoy it. If I worked somewhere else, I’d have to quit and look for a new job every time I wanted to try something new —but not here. AtIBM, we have tremendous chances to try new things we like to do.

Name: Kewei We
Years at IBM: 11
Lives and works in: Beijing, China
Currently working on: Machine Learning Z Systems
Favorite programming language: PL/X
His top 3 beach books for not-so-light reading: “Three Kingdoms” by Luo Guanzhong, “Blood Remuneration Law” by Li Qiang, and “The Shortest History of Europe” by John Hirst

Dinesh Nirmal,  

Vice President, Analytics Development
Follow me on Twitter @DineshNirmalIBM