8. 8
Big Data Proof of Concept (PoC)
The RCG Big Data Proof of Concept
demonstrates the business value of Big Data
using your data in RCG’s Big Data Lab with Big
Data technologies and analytics. This requires
no investment in Big Data hardware, software,
or skills in your IT or business units.
20. Our Brand Promise
Our reputation is built upon the premise that
we are a company that listens.
We bring a creative view to your
business initiative.
We are collaborative and accountable as
we jointly create your solution.
We continuously innovate from concept to result and
help you affect business change.
There will be no surprises.
Ideas. Realized.®
Hinweis der Redaktion
Forrester’s definition of Big Data: “the practices and technology that close the gap between [all types of] data available and the ability to turn that data into business insight.”
12 nodes of Hadoop or NoSQL configuration – this reflects the use of the lab for Proof of Concepts, not necessarily production-level support
½ terabyte of memory
144 terabytes of storage – this provides for a meaningful amount of data to be stored for data science analytics
‘R’ and SAS statistical analysis technologies
Apache Hadoop project software – including HDFS, HBase, Hive, Pig, Sqoop, Yarn, Zookeeper, Mahout, Tez, Flume, Ambari, Oozie, Falcon, Knox, Accumulo, Storm, Kafka, add-ons and connectors to Microsoft, Oracle, Teradata, Informatica, and Talend, and Cloudera, Hortonworks, and MapR Hadoop packages
NoSQL and NewSQL options, including Cassandra, Couchbase, MongoDB, and HPCC
Here are my thoughts on Big Data PoC proposals. I suggest that:
Identify a Business Problem Area be a half day Solution Build type of activity; it may be helpful if this were "free" (no cost for the activity, but build the cost into the costs of the next steps)
Collect Data may require RCG assistance onsite at the rates Rob quoted; this step may take time and should be T&M and not count against a Lab timebox
Load Data into the Lab is when a period of time starts; this will be the Big Data Environment Specialist configuring the environment for the PoC, which can be done while Collect Data is happening, and loading client data into the Lab
Apply Analytics is where the work is; three weeks should be a good start, as long as we can coordinate our analytic resources; it may be desirable to include a Manila resource or two to generate more models and insights
Produce Results and Insights should happen in the third week or so, allowing for an iteration or two with the client
This is one week of a Big Data Environment Specialist ($7,000), three weeks of a Big Data Scientist ($24,000), and 3 weeks for two Manila-based Big Data Analysts (around $12,000), totaling about $45,000 if the costs for step 1 are included. But it will depend on the expectations of the client and how sophisticated the statistical models need to be to meet the expectation.
So, I suggest the proposal to JC Penney should be: We will come in to Identify a Business Problem Area JC Penney wants us to attack, the data needed to analyze it, decide whether we need to help Collect Data, and determine how much Apply Analytics we need to do. We do this for "free" and adjust the price of the PoC depending what expectations JC Penney has for this.
We can say that a ballpark price for a PoC. is $45,000, but that the price can vary based on how extensive the PoC is.
Just some thoughts on the matter . . .