Click here to close now.


Microservices Expo Authors: Yeshim Deniz, Jason Bloomberg, Carmen Gonzalez, Elizabeth White, Liz McMillan

Related Topics: @CloudExpo, Java IoT, Microservices Expo, Open Source Cloud, Agile Computing, Apache

@CloudExpo: Article

The Cure for the Common Cloud-Based Big Data Initiative

Understanding how to work with Big Data

There is no doubt that Big Data holds infinite promise for a range of industries. Better visibility into data across various sources enables everything from insight into saving electricity to agricultural yield to placement of ads on Google. But when it comes to deriving value from data, no industry has been doing it as long or with as much rigor as clinical researchers.

Unlike other markets that are delving into Big Data for the first time and don't know where to begin, drug and device developers have spent years refining complex processes for asking very specific questions with clear purposes and goals. Whether using data for designing an effective and safe treatment for cholesterol, or collecting and mining data to understand proper dosage of cancer drugs, life sciences has had to dot every "i" and cross every "t" in order to keep people safe and for new therapies to pass muster with the FDA. Other industries are now marveling at a new ability to uncover information about efficiencies and cost savings, but - with less than rigorous processes in place - they are often shooting in the dark or only scratching the surface of what Big Data offers.

Drug developers today are standing on the shoulders of those who created, tested and secured FDA approval for treatments involving millions of data points (for one drug alone!) without the luxury of the cloud or sophisticated analytics systems. These systems have the potential to make the best data-driven industry even better. This article will outline key lessons and real-world examples of what other industries can and should learn from life sciences when it comes to understanding how to work with Big Data.

What Questions to Ask, What Data to Collect
In order to gain valuable insights from Big Data, there are two absolute requirements that must be met - understanding both what questions to ask and what data to collect. These two components are symbiotic, and understanding both fully is difficult, requiring both domain expertise and practical experience.

In order to know what data to collect, you first must know the types of questions that you're going to want to ask - often an enigma. With the appropriate planning and experience-based guesses, you can often make educated assumptions. The trick to collecting data is that you need to collect enough to answer questions, but if you collect too much then you may not be able to distill the specific subset that will answer your questions. Also, explicit or inherent cost can prevent you from collecting all possible data, in which case you need to carefully select which areas to collect data about.

Let's take a look at how this is done in clinical trials. Say you're designing a clinical study that will analyze cancer data. You may not have specific questions when the study is being designed, but it's reasonable to assume that you'll want to collect data related to commonly impacted readings for the type of cancer and whatever body system is affected, so that you have the right information to analyze when it comes time.

You may also want to collect data unrelated to the specific disease that subsequent questions will likely require, such as information on demographics and medications that the patient is taking that are different from the treatment. During the post-study data analysis, questions on these areas often arise, even though the questions aren't initially apparent. Thus clinical researchers have adopted common processes for collecting data on demographics and concomitant medications. Through planning and experience, you can also identify areas that do not need to be collected for each study. For example, if you're studying lung cancer, collecting cognitive function data is probably unrelated.

How can other industries anticipate what questions to ask, as is done in life sciences? Well, determine a predefined set of questions that are directly related to the goal of the data analysis. Since you will not know all of the questions until after the data collection have started, it's important to 1) know the domain, and 2) collect any data you'll need to answer the likely questions that could come up.

Also, clinical researchers have learned that questions can be discovered automatically. There are data mining techniques that can uncover statistically significant connections, which in effect are raising questions that can be explored in more detail afterwards. An analysis can be planned before data is collected, but not actually be run until afterwards (or potentially during), if the appropriate data is collected.

One other area that has proven to be extremely important to collect is metadata, or data about the data - such as, when it was collected, where it was collected, what instrumentation was used in the process and what calibration information was available. All of this information can be utilized later on to answer a lot of potentially important questions. Maybe there was a specific instrument that was incorrectly configured and all the resulting data that it recorded is invalid. If you're running an ad network, maybe there's a specific web site where your ads are run that are gaming the system trying to get you to pay more. If you're running a minor league team, maybe there's a specific referee that's biased, which you can address for subsequent games. Or, if you're plotting oil reserves in the Gulf of Mexico, maybe there are certain exploratory vessels that are taking advantage of you. In all of these cases, without the appropriate metadata, it'd be impossible to know where real problems reside.

Identifying Touch Points to Be Reviewed Along the Way
There are ways to specify which types of analysis can be performed, even while data is being collected, that can affect either how data will continue to be collected or the outcome as a whole.

For example, some clinical studies run what's called interim analysis while the study is in progress. These interim analyses are planned, and the various courses that can be used afterwards are well defined, but the results afterward are statistically usable. This is called an adaptive clinical trial, and there are a lot of studies that are being performed to determine more effective and useful ways that these can be done in the future. The most important aspect of these is preventing biases, and this is something that has been well understood and tested by the pharmaceutical community over the past several decades. Simply understanding what's happening during the course of a trial, or how it affects the desired outcome, can actually bias the results.

The other key factor is that the touch points are accessible to everybody who needs the data. For example, if you have a person in the field, then it's important to have him or her access the data in a format that's easily consumable to them - maybe through an iPad or an existing intranet portal. Similarly, if you have an executive that needs to understand something at a high level, then getting it to them in an easily consumable executive dashboard is extremely important.

As the life sciences industry has learned, if the distribution channels of the analytics aren't seamless and frictionless, then they won't be utilized to their fullest extent. This is where cloud-based analytics become exceptionally powerful - the cloud makes it much easier to integrate analytics into every user's day. Once each user gets the exact information they need, effortlessly, they can then do their job better and the entire organization will work better - regardless of how and why the tools are being used.

Augmenting Human Intuition
Think about the different types of tools that people use on a daily basis. People use wrenches to help turn screws, cars to get to places faster and word processers to write. Sure, we can use our hands or walk, but we're much more efficient and better when we can use tools.

Cloud-based analytics is a tool that enables everybody in an organization to perform more efficiently and effectively. The first example of this type of augmentation in the life sciences industry is alerting. A user tells the computer what they want to see, and then the computer alerts them via email or text message when the situation arises. Users can set rules for the data it wants to see, and then the tools keep on the lookout to notify the user when the data they are looking for becomes available.

Another area the pharmaceutical industry has thoroughly explored is data-driven collaboration techniques. In the clinical trial process, there are many different groups of users: those who are physically collecting the data (investigators), others who are reviewing it to make sure that it's clean (data managers), and also people who are stuck in the middle (clinical monitors). Of course there are many other types of users, but this is just a subset to illustrate the point. These different groups of users all serve a particular purpose relating to the overall collection of data and success of the study. When the data looks problematic or unclean, the data managers will flag it for review, which the clinical monitors can act on.

What's unique about the way that life sciences deals with this is that they've set up complex systems and rules to make sure that the whole system runs well. The tools associated around these processes help augment human intuition through alerting, automated dissemination and automatic feedback. The questions aren't necessarily known at the beginning of a trial, but as the data is collected, new questions evolve and the tools and processes in place are built to handle the changing landscape.

No matter what the purpose of Big Data analytics, any organization can benefit from the mindset of cloud-based analytics as a tool that needs to consistently be adjusted and refined to meet the needs of users.

Ongoing Challenges of Big Data Analytics
Given this history with data, one would expect that drug and device developers would be light years ahead when it comes to leveraging Big Data technologies - especially given that the collection and analytics of clinical data is often a matter of life and death. But while they have much more experience with data, the truth is that life sciences organizations are just now starting to integrate analytics technologies that will enable them to work with that data in new, more efficient ways - no longer involving billions of dollars a year, countless statisticians, archaic methods, and, if we're being honest, brute force. As new technology becomes available, the industry will continue to become more and more seamless. In the meantime, other industries looking to wrap their heads around the Big Data challenge should look to life sciences as the starting point for best practices in understanding how and when to ask the right questions, monitoring data along the way and selecting tools that improve the user experience.

More Stories By Rick Morrison

Rick Morrison is CEO and co-founder of Comprehend Systems. Prior to Comprehend Systems, he was the Chief Technology Officer of an Internet-based data aggregator, where he was responsible for product development and operations. Prior to that, he was at Integrated Clinical Systems, where he led the design and implementation of several major new features. He also proposed and led a major infrastructure redesign, and introduced new, streamlined development processes. Rick holds a BS in Computer Science from Carnegie Mellon University in Pittsburgh, Pennsylvania.

Comments (0)

Share your thoughts on this story.

Add your comment
You must be signed in to add a comment. Sign-in | Register

In accordance with our Comment Policy, we encourage comments that are on topic, relevant and to-the-point. We will remove comments that include profanity, personal attacks, racial slurs, threats of violence, or other inappropriate material that violates our Terms and Conditions, and will block users who make repeated violations. We ask all readers to expect diversity of opinion and to treat one another with dignity and respect.

@MicroservicesExpo Stories
DevOps has often been described in terms of CAMS: Culture, Automation, Measuring, Sharing. While we’ve seen a lot of focus on the “A” and even on the “M”, there are very few examples of why the “C" is equally important in the DevOps equation. In her session at @DevOps Summit, Lori MacVittie, of F5 Networks, will explore HTTP/1 and HTTP/2 along with Microservices to illustrate why a collaborative culture between Dev, Ops, and the Network is critical to ensuring success.
Overgrown applications have given way to modular applications, driven by the need to break larger problems into smaller problems. Similarly large monolithic development processes have been forced to be broken into smaller agile development cycles. Looking at trends in software development, microservices architectures meet the same demands. Additional benefits of microservices architectures are compartmentalization and a limited impact of service failure versus a complete software malfunction....
Despite all the talk about public cloud services and DevOps, you would think the move to cloud for enterprises is clear and simple. But in a survey of almost 1,600 IT decision makers across the USA and Europe, the state of the cloud in enterprise today is still fraught with considerable frustration. The business case for apps in the real world cloud is hybrid, bimodal, multi-platform, and difficult. Download this report commissioned by NTT Communications to see the insightful findings – registra...
Achim Weiss is Chief Executive Officer and co-founder of ProfitBricks. In 1995, he broke off his studies to co-found the web hosting company "Schlund+Partner." The company "Schlund+Partner" later became the 1&1 web hosting product line. From 1995 to 2008, he was the technical director for several important projects: the largest web hosting platform in the world, the second largest DSL platform, a video on-demand delivery network, the largest eMail backend in Europe, and a universal billing syste...
Docker is hot. However, as Docker container use spreads into more mature production pipelines, there can be issues about control of Docker images to ensure they are production-ready. Is a promotion-based model appropriate to control and track the flow of Docker images from development to production? In his session at DevOps Summit, Fred Simon, Co-founder and Chief Architect of JFrog, will demonstrate how to implement a promotion model for Docker images using a binary repository, and then show h...
Manufacturing has widely adopted standardized and automated processes to create designs, build them, and maintain them through their life cycle. However, many modern manufacturing systems go beyond mechanized workflows to introduce empowered workers, flexible collaboration, and rapid iteration. Such behaviors also characterize open source software development and are at the heart of DevOps culture, processes, and tooling.
Containers have changed the mind of IT in DevOps. They enable developers to work with dev, test, stage and production environments identically. Containers provide the right abstraction for microservices and many cloud platforms have integrated them into deployment pipelines. DevOps and Containers together help companies to achieve their business goals faster and more effectively.
Containers are revolutionizing the way we deploy and maintain our infrastructures, but monitoring and troubleshooting in a containerized environment can still be painful and impractical. Understanding even basic resource usage is difficult - let alone tracking network connections or malicious activity. In his session at DevOps Summit, Gianluca Borello, Sr. Software Engineer at Sysdig, will cover the current state of the art for container monitoring and visibility, including pros / cons and li...
DevOps Summit at Cloud Expo 2014 Silicon Valley was a terrific event for us. The Qubell booth was crowded on all three days. We ran demos every 30 minutes with folks lining up to get a seat and usually standing around. It was great to meet and talk to over 500 people! My keynote was well received and so was Stan's joint presentation with RingCentral on Devops for BigData. I also participated in two Power Panels – ‘Women in Technology’ and ‘Why DevOps Is Even More Important than You Think,’ both ...
In a report titled “Forecast Analysis: Enterprise Application Software, Worldwide, 2Q15 Update,” Gartner analysts highlighted the increasing trend of application modernization among enterprises. According to a recent survey, 45% of respondents stated that modernization of installed on-premises core enterprise applications is one of the top five priorities. Gartner also predicted that by 2020, 75% of
As the world moves towards more DevOps and microservices, application deployment to the cloud ought to become a lot simpler. The microservices architecture, which is the basis of many new age distributed systems such as OpenStack, NetFlix and so on, is at the heart of Cloud Foundry - a complete developer-oriented Platform as a Service (PaaS) that is IaaS agnostic and supports vCloud, OpenStack and AWS. In his session at 17th Cloud Expo, Raghavan "Rags" Srinivas, an Architect/Developer Evangeli...
The web app is agile. The REST API is agile. The testing and planning are agile. But alas, data infrastructures certainly are not. Once an application matures, changing the shape or indexing scheme of data often forces at best a top down planning exercise and at worst includes schema changes that force downtime. The time has come for a new approach that fundamentally advances the agility of distributed data infrastructures. Come learn about a new solution to the problems faced by software organ...
Our guest on the podcast this week is Jason Bloomberg, President at Intellyx. When we build services we want them to be lightweight, stateless and scalable while doing one thing really well. In today's cloud world, we're revisiting what to takes to make a good service in the first place. Listen in to learn why following "the book" doesn't necessarily mean that you're solving key business problems.
For it to be SOA – let alone SOA done right – we need to pin down just what "SOA done wrong" might be. First-generation SOA with Web Services and ESBs, perhaps? But then there's second-generation, REST-based SOA. More lightweight and cloud-friendly, but many REST-based SOA practices predate the microservices wave. Today, microservices and containers go hand in hand – only the details of "container-oriented architecture" are largely on the drawing board – and are not likely to look much like S...
In their session at DevOps Summit, Asaf Yigal, co-founder and the VP of Product at, and Tomer Levy, co-founder and CEO of, will explore the entire process that they have undergone – through research, benchmarking, implementation, optimization, and customer success – in developing a processing engine that can handle petabytes of data. They will also discuss the requirements of such an engine in terms of scalability, resilience, security, and availability along with how the archi...
Any Ops team trying to support a company in today’s cloud-connected world knows that a new way of thinking is required – one just as dramatic than the shift from Ops to DevOps. The diversity of modern operations requires teams to focus their impact on breadth vs. depth. In his session at DevOps Summit, Adam Serediuk, Director of Operations at xMatters, Inc., will discuss the strategic requirements of evolving from Ops to DevOps, and why modern Operations has begun leveraging the “NoOps” approa...
The last decade was about virtual machines, but the next one is about containers. Containers enable a service to run on any host at any time. Traditional tools are starting to show cracks because they were not designed for this level of application portability. Now is the time to look at new ways to deploy and manage applications at scale. In his session at @DevOpsSummit, Brian “Redbeard” Harrington, a principal architect at CoreOS, will examine how CoreOS helps teams run in production. Attende...
With containerization using Docker, the orchestration of containers using Kubernetes, the self-service model for provisioning your projects and applications and the workflows we built in OpenShift is the best in class Platform as a Service that enables introducing DevOps into your organization with ease. In his session at DevOps Summit, Veer Muchandi, PaaS evangelist with RedHat, will provide a deep dive overview of OpenShift v3 and demonstrate how it helps with DevOps.
All we need to do is have our teams self-organize, and behold! Emergent design and/or architecture springs up out of the nothingness! If only it were that easy, right? I follow in the footsteps of so many people who have long wondered at the meanings of such simple words, as though they were dogma from on high. Emerge? Self-organizing? Profound, to be sure. But what do we really make of this sentence?
Application availability is not just the measure of “being up”. Many apps can claim that status. Technically they are running and responding to requests, but at a rate which users would certainly interpret as being down. That’s because excessive load times can (and will be) interpreted as “not available.” That’s why it’s important to view ensuring application availability as requiring attention to all its composite parts: scalability, performance, and security.