Last week I attended a three-day Gartner Summit in London. For all those who didn’t manage a trip, I want to share with you the biggest insights I gathered from the conference and review those things that impressed me the most. <h2>Business Vibe</h2> Visualization and interaction go hand in hand in with modern businesses. Regardless of industry, the vibe, energy, and atmosphere defines a brand. And at the Gartner Summit, that vibe was entertainment with a dash of class. Organizationally, the event created the ambience of a large music festival from Day 1, with more information than one could possibly absorb at once. From the start, the options were plenty, with alternative presentations taking place concurrently, forcing attendees to plan their agenda in advance, picking entertainment for each slot from a variety of venues. All of this was easier done than said, as Gartner provided a platform where one could download all slides. Such useful forethought allowed participants to follow slides and make notes on personal devices for those presentations that were most interesting. With a litany of available information, I’ll share what I learned from the presentations attended.   <blockquote class="twitter-tweet" data-lang="en"> <p dir="ltr" lang="en">The Gartner Analytics conference kicks off with a focus on Stories, Data, Privacy, and AI. <a href="https://twitter.com/hashtag/GartnerDA?src=hash&amp;ref_src=twsrc%5Etfw">#GartnerDA</a> <a href="https://t.co/EmDOm1gAWO">pic.twitter.com/EmDOm1gAWO</a></p> — Scott Holden (@scottiholden) <a href="https://twitter.com/scottiholden/status/1107631614182543360?ref_src=twsrc%5Etfw">March 18, 2019</a></blockquote> <script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script> <h2>Day 1 The Foundation of Master Data Management - Simon Walker, Sr Principal Analyst, Gartner</h2> There is no time like the present which is why I began the conference with a warm-up by deep diving into Master Data Management, or MDM to those in the know. I will use this opportunity to explain the primary difference is where the source data originates. <ul><li>       Data warehouse sources are transactional or operational systems where the master data simply comes along for the ride after a transaction takes place or an event takes place.</li><li>       MDM Typically incorporate any and all possible Master data sources not just those that are associated with internal systems but with external systems.</li></ul> Another key difference is that MDM focuses specifically on providing an enterprise with a single and consistent view of the key entities within a business by providing their best data representations. Conversely, a data warehouse maintains a full history of changes to specific entities. <h3></h3> <h2> The Foundation and Future of Data and Analytics Governance - Saul Judah, VP Analyst, Gartner</h2> Having had my fill of data management, I switched gears and listened in on a keynote speaker from Gartner who explained the three main topics for this event: <ol><li>      AI</li><li>      Privacy</li><li>      Being data-driven</li></ol> It proved an interesting talk with many industry facts and benchmarks. <img class="wp-image-1829 size-full" src="https://webflow-prod-assets.s3.amazonaws.com/6525256482c9e9a06c7a9d3c%2F65b02339feb7b3e305d96b47_D18girXW0AAul19.webp" alt="Data Science Expierience" width="1200" height="900" /> via Michelle Genser <a href="https://twitter.com/mgenser">@mgenser</a> tweeter <img class="aligncenter size-full wp-image-1804" src="https://webflow-prod-assets.s3.amazonaws.com/6525256482c9e9a06c7a9d3c%2F65b023393dff82577ed34b2c_UNADJUSTEDNONRAW_thumb_69d-1.webp" alt="The Foundation and Future of Data and Analytics Governance" width="1024" height="768" /> <script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script> <h2>Guest Keynote: How to be Human in the Age of the Machine - Hannah Fry</h2> Next was the keynote from Hannah Fry, a lecture full of interesting facts. It was here that an interesting question was posed: Are there hidden stories in our data? Assuming so, how can those hidden stories be used against us? Being human in this age of machines is trying at best. Beyond the idea of pattern recognition, the superficial correlations that are conjured most often alongside talk of “machines”, we have to consider learning more about ourselves and how that can be used to make the world a better place. <h2>Keynote: Magic Quadrant Power Session: Insights on the Markets</h2> Moving away from the human element and refocusing on the age of machines, the next keynote explored magic quadrants. Magic quadrants is a must see. Quadrants and hype cycle are the quintessence of Gartner; Always on point, always up-to-date. At this event I watched a few demos from Vendors (Sisense, Qlick, Rapidminer). It is noteworthy that 90% of the vendors were BI and BI is working hard to catch up with ML, mostly through AutoML. Another large trend here is Augmented Analytic, which tangentially supports BI by catching hard-to-spot insights from the data itself. <img class="aligncenter size-full wp-image-1806" src="https://webflow-prod-assets.s3.amazonaws.com/6525256482c9e9a06c7a9d3c%2F65b0233adde585f524539df4_magicq.webp" alt="Magic Quadrant Power Session" width="1024" height="768" /> <h2>The Foundation of Data Science and Machine Learning: Achieving Advanced Insights and AI for Analytics -  Peter Krensky, Principal Analyst, Gartner</h2> The presentation from Peter Krensky about AI/ML for enterprise proved very interesting. I am happy to report that at our core, what we say to our customers overlaps the sentiments from Gartner. Still, we can learn from this presentation and improve our communication to be more business focused. This session offered high-level introductions to machine learning and data science, specifically the function served by both in a data-driven organization. It covered ideas such as: <ul><li>        Hype versus reality</li><li>        Key trends</li><li>        Proven-use cases</li><li>        Leading technologies</li></ul> More specifically, it answered questions about the manner in which data science and machine learning fit within the analytics and AI strategies of an organization, what early steps data and analytics leaders could take so as to better invest in machine learning/data science, and what the first two years of said implementation should look like for businesses. <blockquote>At the end of the day I was faced with a difficult decision: To go and listen to more Tech/Biz presentations or to see two interesting presentations about helping the public and communities. I opted for the latter, in an attempt to better reflect the commitment to our company purpose: focusing on the most pressing world problems & raising awareness of those problems. In retrospect, it was the right decision, as slides from the technical presentations that were missed are much easier to follow after the fact, but information from the two presentations I attended would have proved more challenging post-presentation.</blockquote> <h2>Case Study: Data for Good — Using Automated Machine Learning to Improve Water Access in Developing Nations. Brian Banks, Director of Strategic Initiatives, Global Environment and Technology Foundation</h2> It was here that the true power of data in its ability to rectify world problems was made clear. This lecture focused on the use of transforming society by way of technical resources and fully leveraging the data available there in by both governments and nonprofit organizations. During this session the presenters reviewed use of an automated machine learning platform that empowered citizen data scientist to change an otherwise disastrous water point data situation into more actionable insights that could be used not just by the data scientist but by their governments and their partners. The presenter drew our attention to an NGO currently helping to provide drinkable water in rural regions of both Africa and Asia. This organization wanted a model and in fact worked for two years trying to build one but they just didn't have the skills or the knowledge necessary to create one. In the end they used AutoML, via DataRobot.  I spoke with Brian and offer our help with the mobile dashboards given the fact that most people in Africa have smartphones but don't have personal computers. Appsilon skills in building such tools and taking care of scaling and adoption can result in helping more and more people worldwide. The first date ended with an evening reception and after so much intense information I was back in a hotel by 9 p.m. and fast asleep. <h2>Day 2</h2> I spent the first half of the second day attending Tech/Biz presentations about: <ul><li>       making model explainable</li><li>       setting the right KPIs to measure</li><li>       operationalization of the models</li><li>       real-time systems (or like business calls it: continuous intelligence)</li></ul> All of these presentations prove useful and still represented only a portion of the available information at the conference. One of the key ideas disseminated to attendees like myself with how to easily explain data science topics to those who are more business-minded. Below are some of the best slides from those morning presentations: <img class="aligncenter size-full wp-image-1807" src="https://webflow-prod-assets.s3.amazonaws.com/6525256482c9e9a06c7a9d3c%2F65b0233b6d43487c68451eb7_performance-pitfalls.webp" alt="" width="1024" height="768" /> Interestingly, data and analytics is starting to be measured not by research but by results. And this reflects our concerns about setting KPIs and ROI expectations up front. <img class="aligncenter wp-image-1808 size-full" src="https://webflow-prod-assets.s3.amazonaws.com/6525256482c9e9a06c7a9d3c%2F65b0233d8587215be0b218fc_IMG_5693-e1553008329674.webp" alt="" width="1168" height="704" /> Many models are not utilized at all, as evidenced below! <h2>Teradata: Using AI for Real-Time Monitoring and Prescriptive Insights - Stephen Brobst, CTO, TERADATA & Stefan Meiler, VP Data Governance & Analytical Services, Siemens Healthineers</h2> The use of artificial intelligence to not only predict but prescribed has been used across multiple industries. The Siemens Healthineers team worked with Teradata team to deploy an advanced analytics which helped to predict necessary maintenance, figure out the usage of spare parts, and improve customer satisfaction. Teradata had a specific session that focused on the reasons why some projects failed. Uncovered where problem such as expectations being set too high, no route to production, no available data, no ROI, or politics. Similar presentations from Gartner had overlapping conclusions. What stood out the most was the bias challenges being mentioned. <a href="https://towardsdatascience.com/what-is-ai-bias-6606a3bcb814">Bias in IT</a> is well described by the <a href="https://www.theguardian.com/technology/2018/oct/10/amazon-hiring-ai-gender-bias-recruiting-engine">case of Amazon AI recruiting tool that favored men</a> or tool used by the court systems that <a href="https://medium.com/@SeoJaeDuk/archived-post-bias-in-ai-85a82ca95c6d">bias against towards black people</a>. <blockquote class="twitter-tweet" data-lang="en"> <p dir="ltr" lang="en">Even at a data and analytics show there can be fun! <a href="https://twitter.com/hashtag/GartnerDA?src=hash&amp;ref_src=twsrc%5Etfw">#GartnerDA</a> <a href="https://t.co/0J7yqdfKoO">pic.twitter.com/0J7yqdfKoO</a></p> — Michelle Genser (@mgenser) <a href="https://twitter.com/mgenser/status/1107632270226178048?ref_src=twsrc%5Etfw">March 18, 2019</a></blockquote> <script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script> <img class="aligncenter size-full wp-image-1810" src="https://webflow-prod-assets.s3.amazonaws.com/6525256482c9e9a06c7a9d3c%2F65b0233f0f3e0214b075de26_pitfalls2-e1553008298501.webp" alt="" width="936" height="672" /> The afternoon was full of great guest sessions presenting case studies. I was especially pleased with the interesting presentation from Lloyds by Nick Blewden & Ali Wasseem. They were incredibly happy about setting up an analytics team using QlickSense. Their dashboards were clean and usable and they didn't have a lot of ML/AI efforts but they do hope to develop models in the future. An example of their work can be found here:<a href="https://cityriskindex.lloyds.com/explore/"> https://cityriskindex.lloyds.com/explore/</a> <img class="aligncenter size-full wp-image-1811" src="https://webflow-prod-assets.s3.amazonaws.com/6525256482c9e9a06c7a9d3c%2F65b0234122f94709a7c09805_UNADJUSTEDNONRAW_thumb_69e-1.webp" alt="use cases" width="1024" height="768" /> <h2>Keynote: Willful Disruption: Seven Digital Disruptions You Might Not See Coming - Daryl Plummer, Distinguished VP Analyst, Gartner</h2> I turned my attention toward the business end near the close of the day. It was here that the keynote speaker from Gartner explained how digital projects move from the optimization level to the transformation level but during that time the ability to disrupt becomes a critical discipline. The speaker noted that coping with these disruptions requires people to recognize, prioritize, and respond appropriately. To that end the presentation examined how companies can evaluate, track, and plan for seven of the biggest disruptions using the digital disruption tool kit provided by Gartner. These 7 digital disruptions are not what you might see coming. Not only was this interesting but they put on a great show. All of those disruptions are much further in the future and can be used for more strategic, long-term planning. <h2>Day 3</h2> <h2>Decision Modeling: Bridging Analytics and Business Processes for Optimum Business Outcomes - Erick Brethenoux, Sr Director Analyst, Gartner</h2> The third day started off bright and early with a business based presentation that focused on faster decisions which are becoming increasingly more critical. To that end business models are relying more heavily upon an increasing amount of data to make those faster decisions but in so doing they are becoming moving targets. And as we move forward that gap is getting bigger, the gap between the amount of continuous availability of data and the decisions that organizations have to execute therein. That moment in between the data and the action is where decisions live and where organizations need to use a combination of AI techniques in order to maximize their outcomes. This decision modeling presentation focus heavily on the space where analytical modeling meets with process design. The overall goal was to rethink the company processes by taking into account decision support systems and automation. <h2>Top Technology Trends in Data and Analytics That Will Change Your Business - Gareth Herschel, VP Analyst, Gartner & Donald Feinberg, Distinguished VP Analyst, Gartner</h2> Technology and business converged at the next presentation. With this top technology trends presentation we went beyond AI as a driver of rapid change in data and analytics. For the next generation using augmented analytics tools, there will be changes in the use of continuous intelligence, in the interpretations of data streams from IoT, and in how and where and analysis can be deployed. During this presentation the potential business impact of all of this was explained as it allows organizations to better prioritize which innovations they will enact to drive business digitally. While the talk wasn't a keynote it was still a great presentation. It was much more down-to-earth compared to the aforementioned keynote speeches and at the same time it proved more actionable. I have no doubt that we will use the knowledge from this presentation. Personally, I'm always overjoyed when Gartner confirms what I think about block chain, and what I have thought for quite some time. &nbsp; <h2>Birst, an Infor Company: CDO Talk — Data Strategy in the Cauldron of Business as Usual - Peter Jackson, Group Director Data Sciences, Legal and General</h2> The last of the presentations, bringing to a close an eventful and informative conference was a business based discussion presented by the new Chief data officer about data strategies that provide businesses with better value and more positive outcomes. The presenter provided to real examples of strategies and technologies that he has selected and used these examples to explain how his data vision was developed and executed. Co-author of The Chief Data Officer’s Playbook, a must read! <h2>Networking Experience</h2> Networking at this conference falls into the “worked” and “didn’t work” categories at the same time. It was easy and hard. What made it easy? Almost everyone I talked to was incredibly friendly and had something interesting to say. What made it hard? Starting conversations with strangers is always hard as it forces you to step out of your comfort zone. This conference was a bit harder for networking as there were many groups in attendance, specifically people who came from the same company and on the other end of the spectrum There were single people who attended on their own and spent most of the time looking at their phones. Either situation can make it a bit awkward for the approach. By the end of the conference I ended up connecting with some great people! &nbsp; <h2>Final thought</h2> Overall, the conference provided actionable information on technology and business topics, specifically highlighting areas where they overlapped. The best presentations were those which revealed how our organization is in alignment with Gartner, where similarities manifested and where future direction was provided. All in all, in spite of only being able to attend certain presentations, the information was overwhelmingly good. Long-term this event will remain a positive memory for me and I am very thankful for this opportunity to grow. <blockquote class="twitter-tweet" data-lang="en"> <p dir="ltr" lang="en">Sketchnote of <a href="https://twitter.com/hashtag/GartnerDA?src=hash&amp;ref_src=twsrc%5Etfw">#GartnerDA</a> opening keynote <a href="https://t.co/EklBNMhbFo">pic.twitter.com/EklBNMhbFo</a></p> — John Michl (@JohnMichl) <a href="https://twitter.com/JohnMichl/status/1107645518816165888?ref_src=twsrc%5Etfw">March 18, 2019</a> &nbsp;</blockquote>