What is our primary use case?
The primary use case is for our various financial products. One is a trading platform called MarkitSERV. All the world-leading financial institutes are on-boarded to that client. They use the platform for trading. On the backend, we have the data warehouse of that trading platform, which helps produce quick analyses of the trading data and prepare reports on demand. For that purpose, we are using the OBIEE. We have defined a lot of trends in OBIEE so you can compare year-on-year and month-on-month data. We have various ways you can measure your numbers.
Performance is just okay. It doesn't have in-memory data. For every new analysis we prepare, it will go to the database and query the SQL. That takes a long time. That is the main challenge we are facing. People nowadays are very demanding. They want their analysis to come up very quickly, at the click of a button.
What is most valuable?
The most valuable feature of Oracle OBIEE, for me, would be the Web Catalog. From a development perspective, I do not need any client or any software to be installed on my laptop or desktop. Everything is browser-based, so I just open the browser and start development.
Similarly, for the end-user, it provides self-analytics services where the end user can drag and drop various fields, facts, and dimensions and prepare their reports. They can have them in the form of charts, etc. Then they can save them in the Web Catalog.
I have not seen this Web Catalog in any other tool. Web Catalog is kind of like Windows Explorer, where you can define your own folder and whatever analysis you are doing and you can save the analysis for future reference.
What needs improvement?
One feature I would definitely like to see is the ability to provide the in-memory data. Oracle might have some plugins but, as of now, that feature isn't available out-of-the-box. You might have to purchase that feature. This is the feature that is most lacking, which we would like to have in the next release.
Also, the overall architecture. Currently, it comes in an Oracle suite. We want to have it come as a separate product and which is very easy to install, migrate, deploy etc.
For how long have I used the solution?
Three to five years.
What do I think about the stability of the solution?
I wouldn't say it's that stable. We are using it in production and we used to have one failure per month. When we're not able to figure out the reason, we just restart the services to sort it out. That might be an environmental issue, we're not sure. It could also be a tool issue. When we approached Oracle support, we could not find out exactly what caused the instability.
What do I think about the scalability of the solution?
It's fairly scalable. It doesn't have its own engine, so everything depends on how much data your database can process.
How is customer service and technical support?
We have used technical support. We didn't find it very useful. The responses are very slow. As the tool is very, very complex, there are a lot of scenarios that we need to go through with them. Most of the time we are not able to simulate the issues with them. Depending on the application-side logs that we provide to support, they start their investigation, but most of the time it is not that useful.
How was the initial setup?
I wasn not involved in the initial setup. When I joined the organization, it was already set up. I would say that we're using only 20 to 25 percent of the features of OBI.
Setup and migration are very complex. Technically, Oracle accessory is extremely complex. These days, the industry demands a kind of Excel analysis, which is very easy to install, set up, and it's very developer-friendly as well as end-user friendly. OBIEE lacks all those features. We had to migrate from 11g to 12c and it took around six months because of the extreme complexity of the tool.
Which other solutions did I evaluate?
We are planning to move away from Oracle OBIEE, and to move to another tool that has in-memory data processing capabilities. Oracle doesn't provide that capability very easily, so we have performance challenges, etc. We are exploring Tableau and QlikView. They provide in-memory calculations and in-memory processing, which results in very fast throughput so that the data updates into the dashboard.
In terms of choosing a vendor, the most important criteria for me is the ability to create near real-time analysis, which OBIEE is lacking at the moment. Another is very good performance, which means it should have its own engine to process the data or even to do some small ETL operations on the data. In addition, it's the simplicity to develop and architect the solution.
What other advice do I have?
All the drawbacks which I mentioned above, I would like to have all of those resolved in OBIEE. First of all, its installation, configuration, and setup should be very easy. It comes with its own application server and that creates a lot of conflict files on the server side. You have to be very careful while configuring all those files.
Technical support is another issue, where we generally do not get adequate support. And on Google, there is a lack of good material for OBIEE 12c. If you compare it with other tools like Tableau and QlikView, they have excellent community sites where you post your question and you'll immediately get a reply.