How has it helped my organization?
Tableau has improved my organization in a variety of ways, one of its uses being that of data analysis. It provides a server platform for sharing information. We use it for internal collaboration, as well as other tools for data catalog, for creating the dashboards, for preparing the data in preparation of creating the dashboards, called an ETL extract, and as a tool to transform and load. Tableau is a platform that has several products, perhaps four or five, that average for the fifteen of big data, data evaluation and data collaboration. No specific aspect can be used for this and it can be employed in marketing and finance. It serves the needs of data analysis and providing an algorithm for machine learning. For instance, you can have a logistic regression to analyze whether a specific customer is a good bet or not, such as a bank that is contemplating the loan of money. It allows you to visualize and analyze your data no matter what it may be, though it can be used for an alternate solution.
What is most valuable?
A feature I have found most valuable is the ease of use and straightforwardness, in addition to the flexibility of Tableau. I like the fact that Tableau can connect to a wide variety of databases, be on cloud or on-premise. Tableau can connect to over 100 database types, including structured and non-structured databases. Tableau can connect to a PDF and extract all the tables you have in that PDF. Suppose you have a one hundred-page PDF containing sixteen tables of data. Tableau can connect to that PDF and extract its data. Tableau can connect to Google Drive, to a host of marketing portals on the internet, to cloud companies such as AWS or Alibaba and to many different types of databases. That's one huge advantage of the tool.
While it can be complex if you need to alter a lot of parameters, it provides simple installation. It is very easy. All you would need to do if you have only one Tableau running server is to employ the maximum connection and install a license column in Adobe Reader.
What needs improvement?
An area needing improvement involves the complexity of the product should you need to alter a lot of parameters.
Definitely speaking, it's straightforward and it's very easy. Implementation problems can be dealt with by the client, in place of the user consultant. Let me give you some examples of things that could take long in a Tableau implementation. Suppose you have five different business areas in your company: marketing, supply chain, finance, HR and procurement. Let us suppose that access to HR salaries is not company-wide but is limited to only a select number of people in HR, such as the manager or the director of the department. Yet, I want people in the supply chain to be able to see and access different data from different areas. While this would not be technically difficult it would be time consuming if the businesses are very particular. There may be many policies involved in access authorization, in data availability and the like.
This can involve a very strict security process using an outside identity provider. Instead of just logging in your username and password, you may have different technologies which are more safe and secure that need different providers to interface in Tableau. Depending on the need, this will be time consuming. For instance, while I don't know how this would be in your country, suppose you have an identity provider, in Brazil, marketing in Tableau. If you go to Asia, you may sometimes have a bio-metric identity that your hand or fingers employ which is going to get back at you. In that circumstance, they are going to send you a number or a code in your cellphone, requiring two steps, one to enter the bank and the other to withdraw your money. So, these things we call an outside identity provider, meaning a different vendor or different companies who manage the servers of managing identities. These would entail an integration with Tableau and these outside companies for security purposes. This would involve them sending me files and me sending them back in order to authenticate the user into the Tableau server.
This can be time-consuming because they involve or require a different partner. Tableau is made for basic needs, such as requiring a user and a password to log in to the server; an unsophisticated architecture; or use of a single instead of a cluster of servers. If you have non-specific data security needs or you just want to analyze and sell data, that can take less than a day. But if you have technical servers, many interfaces, different providers and more serious processes, that will be time consuming.
While Tableau does integrate with Arc server and Python server, the integration process is slow and the information is integrated in a protracted fashion. Sometimes your data will vary. You may have a vector of data. You may have a matrix of data. For some algorithms we do not use regular data, but a different data structure. Tableau does not work with these different data structures. As such, interfacing with Arc server and Python server, which are still languages that are widely used in machine learning, all happen slowly. It does not happen by a matrix of data and data vector.
For how long have I used the solution?
I have been using this solution for five years.
Which solution did I use previously and why did I switch?
In the past I worked with Oracle E-Business Suite while working with ERP markets over a thirteen or fifteen year period. Yet for the past five years I've been focusing mainly on artificial intelligence, machine learning, big data and the use of other software, such as Tableau and Azure for the purpose of developing and building data to create algorithms and visual dashboards to show the data. It's been around five years since I have turned my focus solely to big data and machine learning.
How was the initial setup?
Definitely speaking, the initial setup was straightforward and very easy.
Which other solutions did I evaluate?
Another option I evaluated is Power BI from Microsoft. It's cheaper than other solutions and requires fewer different packages. The major competitor of Tableau is Power BI from Microsoft and Microsoft's much cheaper than Tableau. But Microsoft usually requires me to be on Microsoft cloud Azure. You have to buy other solutions for an integrated solution. At the end your cost will be much higher. So Tableau is more flexible.
In Tableau, I can have a scatter plot with millions of marks. Suppose I have a graph that plots my value against my process and each dot in the graph is a sale that I've made. So I have 30 million dots in this graph reflecting my 30 million sales. Tableau can run this easily and fast. Power BI cannot. Power BI has a limitation of 13,500 marks, meaning Tableau has more capacity in delivering data than its competitors.