Business Intelligence
is process of analyzing the data to produce meaningful and conclusive
information.
Bird eye view of BI enabled system looks something like below –
We are
going to discuss a bit in this article about Visualization.
Choosing
well suited reporting framework for your system is very important decision to
make. In many scenarios people make just in time decisions by doing quick and
dirty research which may become a big issue as project grow.Choice of
tool can be tricky and depends on whether you are planning to migrate existing visualization
framework to newer or choosing it for the first time.
Roadmap of
the tool and if any open source, then number of contributors is something you
should see.
Ask below questions
before you choose the platform.
How much data are we going to
process?
Getting
clear understanding on how much data you are going to process and visualize is
one the key point of investigation before you choose any tool, and the
projection of data size over the time. e.g. tool required for IoT based system
would be different than that of CRM like smaller systems. You also need to know
what type of data source you are going to connect to (Relational or NoSQL).
Different
data connectivity/processing approaches are required based on the data size to
process e.g. if you are going to use Tableau/Logi like platforms then you better
write extracts to process large data, but if you are choosing JavaScript based charts,
you need faster API to get and show the summarized information.
What is our budget?
The very
important question during initial phase of the project. Everything come and
stop here, the budget. You can build reporting for IoT system based on
JavaScript backed up with lightning fast micro-services OR you can ease out
your life by using the ready frameworks like Tableau, Pentaho, Logi etc.
The costing
in both approaches differs a lot, frameworks work in named and core licensing
that cost large bucks whereas other approaches are cost effective; it’s a
tradeoff you may need to do based on the budget you have and of course the time.
Do consider
the cost for various projections for data, number of users, report types etc.
What should be the time to
market?
If you use
low cost solutions, you likely end up spending more time (and hence money over
the time) on delivering new reports, whereas with ready frameworks time to
market may be amazingly less.
What types of reports are
expected?
Usually reports
are estimated based on their complexity such as Easy, Moderate or Complex. Reports
could be as simple as list or may have complex layout with different levels of
grouping, sorting, related sections, connecting to various data sources etc.
Again, time
to market is usually a deciding factor.
Need of
real time data vs. batch mode access is another key requirement to know.
How easily can we achieve
non-functional aspects?
Performance,
Scalability and the Security are key aspects that one must consider while implementing
reporting solution, it is important to know the priority and the allocated
budget to support these aspects.
With enterprise solutions you get most of it
out of the box whereas for custom solutions you need to invest time to
implement these tailor made features.
#Tableau
#Tableau
Tags:
Tableau
0 comments :
Post a Comment