We just concluded Dell EMC World 2017 in Las Vegas, our largest customer and partner event of the year. The entire family of Dell Technologies businesses came together to showcase our shared vision for the future of enterprise technology.
Dell EMC World is a 4 day showcase of ingenuity and innovation focused on the ongoing digital transformation of the world and the information technology that will power it. More than 13,500 IT practitioners, business decision makers, analysts and influencers from 122 countries were there to see our keynote speakers, the solutions on the Expo floor, attend technical breakout sessions, network, and even find some time to enjoy the experience that is Las Vegas.
The Solutions Expo included more than 135 exhibits featuring every member of the Dell Technologies family of businesses. Our hands-on labs allowed attendees the opportunity to experience a live ecosystem of technology they can use to realize their digital future. We held more than 500 breakout sessions across three tracks: Technology, IT Leadership and Code & Modern Operations that gave every attendee the opportunity to walk away with practical information they can use on their very first day back in the office.
With all that was going on at the show you can imagine that the big social media channels were buzzing. The show hashtag, #DellEMCWorld, appeared on the trending list of Twitter for the best part of 3 days. As a Dell EMC marketing professional focused on big data and analytics I was very excited to see that the three keywords used most often along with #DellEMCWorld were:
#Iot #AI #Big Data
This is very consistent with what we heard from the attendees we met with in March at Strata + Hadoop World in San Jose. Our lead product manager for Hadoop solutions, Armando Acosta, and I both attended that show and the most frequently asked questions were related to the evolution of data analytics architectures built on the Hadoop ecosystem including: 1) when and how can we move from batch to real time analytics, 2) is Spark right for me and 3) how can I use machine learning/deep learning with my Hadoop investments.
I'm going to attempt to say something briefly about all of these topics and only ask you for about 15-20 minutes of additional reading time. One way I can do that is by pointing you to some additional resources where it makes sense in the context of this discussion and I will also put some links at the bottom so you can bookmark those for later. Let's go for it.
Artificial Intelligence/Machine Learning/Deep Learning
This is obviously too many very technical topics to cover in a single section of a blog article, but, I’m going to try to make a couple of statements that I think you may find useful in approaching these subjects in the fewest words I can and then point you to a great 12 minute video from Dell EMC World. First, the distinction between AI, machine learning and deep learning can be ignored in most situations without much impact. All three disciplines involve the implementation of a multi-step process that is typically referred to as a data analytics pipeline. These pipelines involve a combination of:
- the execution of packaged statistical functions on large batches of data to discover and train useful classification and/or predictive models
- code you develop yourself for automation and management of data ingestion, model training, model comparison/evaluation, etc.
- incorporation of the best model results into custom applications for adding real-time intelligence based on patterns discovered in the first step
- rinse and repeat to update and improve your classification or predictions over time.
A key requirement of achieving a good implementation of machine learning or deep learning is that the value of the system improves over time as the model “sees/processes” more data with little or even no human supervision. Fun stuff.
These capabilities have to be developed over time in a series of incremental development stages that build up to a more complete implementation. You need a team with many different skills to be successful. You must hire or rent fully trained and experienced data scientists to scope and supervisor the effort - there are no magic tools or short courses that will provide a substitute for training and experience.
Here is a link to a really interesting video that covers a lot of these topics including the use of batch/supervised machine learning, real-time unsupervised analytics for stopping credit card fraud at the time of the transaction and how the Dell EMC and Master Card teams partner to get the most of Cloudera Hadoop, Spark, machine learning together with the right infrastructure. See Tony Parkinson, VP Solutions North America at Dell EMC and Nick Curcuru, VP Big Data Analytics at MasterCard on theCUBE at DellEMC World 2017.
Internet of Things (IoT)
One of the analysts conducting the Dell EMC and Master Card interview remarked that he never previously would have put machine learning and Dell EMC together. We need to work on that. You might also be surprised that the combined capabilities of Dell Technologies has the following solutions for tackling the largest and most sophisticated IoT architectures in the world:
- purpose-built Edge gateways that support analytics at the edge of the network featuring a ruggedized enclosure with a variety of input/output connections
- ruggedized, fanless embedded box PCs for the industrial PC market
- Create a flexible and powerful Internet of Things ecosystem with analytics enabled at the edge, data center core and the cloud.
- Put security first to safely deploy IoT initiatives and achieve functional integrity and data security.
- Scalable storage
- Object-based geo-distributed data lake that has the flexibility, scalability, compliance, and sophisticated architecture to support data on an IoT scale.
- Contributions to open source
- Dell and other industry leaders have joined forces to create EdgeX Foundry, the interop platform for the IoT edge. Learn more about this Linux Foundation collaborative project.
- Turnkey platforms
- Dell EMC Converged Platforms solution consisting of Native Hybrid Cloud and Analytic Insights Module
- Professional Services
- Align the business and IT around a strategic initiative
- Apply data science to your business models and data sources
I spend a fair amount of time sharing information on social media so I realize that everyday the hash tags associated with artificial intelligence, machine learning, deep learning and big data analytics are flooded with hype and misinformation. I see 100 Tweets a day telling me about "easy data science" and the "5 steps to realize value from big data". My message is that there are no just add water and stir solutions - plain and simple. Data science takes hard work and focus just like everything else worth doing. One of the things I like most about working at Dell EMC is that our sales and marketing professionals are dedicated to building long term relationships with our customers. The data that suggests it costs 5-10 times more to acquire a new customer compared to retaining an existing one is proof that paying attention to relationships is what keeps businesses successful in the long run and we plan to be around for a good long time.
In order to help you decide what you can do right now, here is a tip from our Director of Data Analytics for North America, Anthony Dina:
When it comes to organizations and their data analytics journeys, a quote often attributed to Mark Twain seems especially appropriate: “The secret to getting ahead is getting started.” Read more here
Please stop by and talk to us in the expo hall if you are attending Data Works Summit 2017 in San Jose June 13-15, 2017. Bill Schmarzo, CTO - Dell EMC Services (aka “Dean of Big Data”) author of “Big Data: Understanding How Data Powers Big Business” and “Big Data MBA: Driving Business Strategies with Data Science” will be giving a keynote presentation on Wednesday, June 14th.
You can also check out the entire Dell EMC Big Data Portfolio on our website and chat with a customer service representative if you have any questions on how to get started. Thanks for reading and please consider taking us on the journey with you!
Phil Hummel, EMCDSA
|Share via email||Share on Linked In||Share on Twitter|