Gartner Data & Analytics Summit 2018
On March 19-21, the Gartner Data & Analytics Summit conference was held in London. I was a visitor to this event and I want to share with you my thoughts and observations.
I was familiar with the activities of Gartner company long before the conference thanks to the famous “magic quadrants” and “HYIP cycles”. Going to the conference, I formulated the following personal goals:
The expectation that the organizers will be able to surprise me with something new has been formed purely speculative. A company exploring a high-tech and dynamic digital market must have itself demonstrated the best practices of this market during the conference.
')
The first thing I want to mention is the mobile application of the conference - Gartner Events Navigator . At first, looking at the required privileges, I did not want to install it, because I was afraid that I would not get anything valuable for myself (besides what was already on the site), but I would provide a lot of information about myself to a third-party company. But, after reading all the features of the application, I installed it and in the end I was very pleasantly surprised at how popular and thoughtful it was made by the authors. Judge for yourself, in the application you get:
a) Personal planner with the ability to watch the schedule, schedule a visit to your favorite presentations and download materials in PDF to your phone
b) Social network and chat participants
c) Hall layout
d) Feedback Tools
e) Advisory service, which will select the best presentations for the visit, depending on your interests.
I actively used the application for all three days and I think what happened to me was what the experts in the field of big data ethics said recently - users will be happy to share their personal data with the application if they understand what value they get in return.
The second thing I would like to note is the digitization of the human flows at the conference.
Each participant at the conference wore a badge with a bar code, and at the entrance to each hall and at each demonstration booth, the organizers with the help of such a device scanned all visitors who showed interest in this particular topic:
Considering that each participant during registration indicated enough professional information about himself, there were only about 1500 participants, and the event lasted 3 days, this whole Internet ofthings of people generated a very interesting data file, which, besides direct study, you can also monetize.
Plus, I liked it more - each performance began and ended strictly on time, minute by minute.
Turning to the conference itself, it is impossible not to note the supersaturated (pdf) agenda. There were several types of events: presentations, round tables and workshops and an exhibition with more than 50 vendors. It was necessary to accept in advance that it would not be possible to visit all 100% of the events, from the very beginning it was necessary to tune in to a compromise and make a painful choice.
For me, it is interesting to post factum to analyze the statistics of topics covered at the conference (presentations and round tables):
From my point of view, this disproportion in favor of Data Governance is somewhat artificial. It seemed to me that the true leader in terms of user interest was the topic “AI & ML”, in any case, I was not able to attend any round table or master class on ML due to the fact that registering for them by the beginning of the event were closed, while there were free slots at Data Governance events.
Below are several theses / terms / thoughts that I remember after the last three days of presentations and reports:
Organizational structure - the most common hybrid model. There is a Chief Data Officer who reports to the CEO or CIO. Data Scientists in an organization are usually of two types: with a specific business function or in a separate Center of Excellence. The purpose of the Center of Excellence is RnD and innovative tasks that cannot be attributed to any of the existing businesses. CDO is responsible for coordinating Data Scientists. He also defines the goals and development strategy.
Only 20% of the surveyed visitors have a CDO role in their staff structure. Gartner forecast - by 2020 this figure should grow to 40-50%.
Data Lake - the most common implementations are:
1. Hadoop
2. Object Store (for example, Amazon S3)
3. RDBMS
Data Lake is not the “killer” of Enterprise DWH, but complements it in the corporate IT landscape.
Data Lake / Hadoop security is a real risk, both for the reason that there are no standard tools, unlike RDBMS, and for the reason that almost no one invests in the implementation of security policies in Hadoop: authentication, authorization, auditing.
Data Literacy (literacy with tz data management) - an asset in which to invest. Only about 5% of organizations can boast that they have implemented end-to-end programs to improve Data Literacy at all levels - from Executive Management to Junior specialists.
Logical DWH is a way of organizing access to data, in which no centralized data storage takes place, as, for example, in DWH / Data Lake. However, for users, such an organization is completely transparent, and they can safely use the data without thinking about how access is organized. (It turns out that there are already a lot of decisions on this topic, see below.)
Open Source - many vendors are starting to make Community versions of their commercial products and distribute them for free, thus recognizing the viability of such a model, and at the same time earning a loyal fan base (However, not all. I had a dialogue with one vendor about how I I can try to solve them. The answer was such that in order for them to send me the activation keys, I have to write them a letter, sign an NDA, draw up a PoC contract, in which I write success criteria and obligations to purchase their product, if all criteria are and success will be fulfilled. ¯ \ _ (ツ) _ / ¯)
In addition to presentations and reports that went in several streams, an exhibition area operated in the center of the conference in the large hall, where you could get to know the representatives of the companies and see the demo of their products.
In a structured way, here are some companies that managed to see there:
Major well-known vendors:
- Oracle - presented their cloud platform
- Qlik - Self-service BI solutions
- Tableau - Self-service BI solutions
- Ataccama - Data Quality, Master Data Management, Data Govenance Solutions
- Google Cloud - cloud platform
- IBM - introduced solutions for Data Science: DSX & SPSS
- Attunity - data replication / CDC solution
- Teradata
- SAS
- SAP
- Microstrategy
- Informatica
- Information Builders
Hadoop commercial builds:
- Cloudera
- MapR
- Hortonworks
Solutions for Data Science:
- Dataiku
- Rapid Miner
- R Studio
- Angoss
Solutions for Data Governance:
- Alation
- Backoffice Associates
- Collibra
- Semarchy
- Stibo Systems
Self Service BI Solutions:
- Arcadia Data
- Looker
- Sinequa
- ThoughtSpot
Software for building a logical DWH:
- Actian
- BI Builders
- Denodo
- Domo
- Dremio
- Iguazio
- sisense
- Snowflake
- Trifacta
- Wherescape
Of course, not all the solutions that were presented, but to look at them all in such a short time in the intervals between presentations is an almost impossible task, especially if we consider that after the third or fourth company in the head everything starts to mix.
Briefly about the findings
The goals were achieved in general. An idea was received about the direction in which the market is developing, what players are there, and where we are now in terms of technology (Hadoop, Data Lake, Streaming) and processes (Governance, Security, staff training, etc.) regarding what was heard on the summit. These findings will form the basis of the next development plans.
What was most surprising was the large number of relatively young vendors who are aiming at the “Logical DWH” segment, and for which, in general, the giants do not have time. For me, “Logical DWH” is definitely a topic that has to be dealt with in more detail and in depth.
Well, in general, I confirmed for myself the conclusion that such events are useful from the point of view of expanding horizons and understanding in which direction the progressive community is moving, so that it would be easier to understand where to develop further.
I was familiar with the activities of Gartner company long before the conference thanks to the famous “magic quadrants” and “HYIP cycles”. Going to the conference, I formulated the following personal goals:
- discover something new at Gartner itself;
- search for new ideas in the field of work with data;
- well, and a bonus - to learn the art of public speaking from world-class specialists (but this topic is beyond the scope of this article).
The expectation that the organizers will be able to surprise me with something new has been formed purely speculative. A company exploring a high-tech and dynamic digital market must have itself demonstrated the best practices of this market during the conference.
')
The first thing I want to mention is the mobile application of the conference - Gartner Events Navigator . At first, looking at the required privileges, I did not want to install it, because I was afraid that I would not get anything valuable for myself (besides what was already on the site), but I would provide a lot of information about myself to a third-party company. But, after reading all the features of the application, I installed it and in the end I was very pleasantly surprised at how popular and thoughtful it was made by the authors. Judge for yourself, in the application you get:
a) Personal planner with the ability to watch the schedule, schedule a visit to your favorite presentations and download materials in PDF to your phone
b) Social network and chat participants
c) Hall layout
d) Feedback Tools
e) Advisory service, which will select the best presentations for the visit, depending on your interests.
I actively used the application for all three days and I think what happened to me was what the experts in the field of big data ethics said recently - users will be happy to share their personal data with the application if they understand what value they get in return.
The second thing I would like to note is the digitization of the human flows at the conference.
Each participant at the conference wore a badge with a bar code, and at the entrance to each hall and at each demonstration booth, the organizers with the help of such a device scanned all visitors who showed interest in this particular topic:
Considering that each participant during registration indicated enough professional information about himself, there were only about 1500 participants, and the event lasted 3 days, this whole Internet of
Plus, I liked it more - each performance began and ended strictly on time, minute by minute.
Turning to the conference itself, it is impossible not to note the supersaturated (pdf) agenda. There were several types of events: presentations, round tables and workshops and an exhibition with more than 50 vendors. It was necessary to accept in advance that it would not be possible to visit all 100% of the events, from the very beginning it was necessary to tune in to a compromise and make a painful choice.
For me, it is interesting to post factum to analyze the statistics of topics covered at the conference (presentations and round tables):
From my point of view, this disproportion in favor of Data Governance is somewhat artificial. It seemed to me that the true leader in terms of user interest was the topic “AI & ML”, in any case, I was not able to attend any round table or master class on ML due to the fact that registering for them by the beginning of the event were closed, while there were free slots at Data Governance events.
Below are several theses / terms / thoughts that I remember after the last three days of presentations and reports:
Organizational structure - the most common hybrid model. There is a Chief Data Officer who reports to the CEO or CIO. Data Scientists in an organization are usually of two types: with a specific business function or in a separate Center of Excellence. The purpose of the Center of Excellence is RnD and innovative tasks that cannot be attributed to any of the existing businesses. CDO is responsible for coordinating Data Scientists. He also defines the goals and development strategy.
Only 20% of the surveyed visitors have a CDO role in their staff structure. Gartner forecast - by 2020 this figure should grow to 40-50%.
Data Lake - the most common implementations are:
1. Hadoop
2. Object Store (for example, Amazon S3)
3. RDBMS
Data Lake is not the “killer” of Enterprise DWH, but complements it in the corporate IT landscape.
Data Lake / Hadoop security is a real risk, both for the reason that there are no standard tools, unlike RDBMS, and for the reason that almost no one invests in the implementation of security policies in Hadoop: authentication, authorization, auditing.
Data Literacy (literacy with tz data management) - an asset in which to invest. Only about 5% of organizations can boast that they have implemented end-to-end programs to improve Data Literacy at all levels - from Executive Management to Junior specialists.
Logical DWH is a way of organizing access to data, in which no centralized data storage takes place, as, for example, in DWH / Data Lake. However, for users, such an organization is completely transparent, and they can safely use the data without thinking about how access is organized. (It turns out that there are already a lot of decisions on this topic, see below.)
Open Source - many vendors are starting to make Community versions of their commercial products and distribute them for free, thus recognizing the viability of such a model, and at the same time earning a loyal fan base (However, not all. I had a dialogue with one vendor about how I I can try to solve them. The answer was such that in order for them to send me the activation keys, I have to write them a letter, sign an NDA, draw up a PoC contract, in which I write success criteria and obligations to purchase their product, if all criteria are and success will be fulfilled. ¯ \ _ (ツ) _ / ¯)
In addition to presentations and reports that went in several streams, an exhibition area operated in the center of the conference in the large hall, where you could get to know the representatives of the companies and see the demo of their products.
In a structured way, here are some companies that managed to see there:
Major well-known vendors:
- Oracle - presented their cloud platform
- Qlik - Self-service BI solutions
- Tableau - Self-service BI solutions
- Ataccama - Data Quality, Master Data Management, Data Govenance Solutions
- Google Cloud - cloud platform
- IBM - introduced solutions for Data Science: DSX & SPSS
- Attunity - data replication / CDC solution
- Teradata
- SAS
- SAP
- Microstrategy
- Informatica
- Information Builders
Hadoop commercial builds:
- Cloudera
- MapR
- Hortonworks
Solutions for Data Science:
- Dataiku
- Rapid Miner
- R Studio
- Angoss
Solutions for Data Governance:
- Alation
- Backoffice Associates
- Collibra
- Semarchy
- Stibo Systems
Self Service BI Solutions:
- Arcadia Data
- Looker
- Sinequa
- ThoughtSpot
Software for building a logical DWH:
- Actian
- BI Builders
- Denodo
- Domo
- Dremio
- Iguazio
- sisense
- Snowflake
- Trifacta
- Wherescape
Of course, not all the solutions that were presented, but to look at them all in such a short time in the intervals between presentations is an almost impossible task, especially if we consider that after the third or fourth company in the head everything starts to mix.
Briefly about the findings
The goals were achieved in general. An idea was received about the direction in which the market is developing, what players are there, and where we are now in terms of technology (Hadoop, Data Lake, Streaming) and processes (Governance, Security, staff training, etc.) regarding what was heard on the summit. These findings will form the basis of the next development plans.
What was most surprising was the large number of relatively young vendors who are aiming at the “Logical DWH” segment, and for which, in general, the giants do not have time. For me, “Logical DWH” is definitely a topic that has to be dealt with in more detail and in depth.
Well, in general, I confirmed for myself the conclusion that such events are useful from the point of view of expanding horizons and understanding in which direction the progressive community is moving, so that it would be easier to understand where to develop further.
Source: https://habr.com/ru/post/351912/
All Articles