• Used Clickstream analysis to fetch the data into HDFS and then analyze it.• User clicks on a product in the website.• Load Balancer then contacts the webserver to check which server is available.• The available webserver then sends the data to log4j.• Log4j send the data to Flume appender and stores the data in a log4j file and also in a database.• With the help of flume appender, the data is sent to HDFS through source , channel and sink.• Here the source is the website, channel is avro server and sink is HDFS.• With the help of Sqoop the data is imported from HDFS and into MySql.• Another table with the same schema like in MySql is created in Hive and Pig.• Data is then loaded in Pig as well as Hive table.• With the help of Pig, the data is sanitized to clear any duplication.• Finally with the help of Hive terabytes and petabytes of data are processed.• Click data flow and user data flow are the two tables used here where URL is the primary key in click data flow table and user id is the primary key in user data flow.

Monitoring Payments Batch processing like BACS, CHAPS, STP, SWIFT, CAML jobs. Mainframe IBM Z-Series, SQL, Cloud Operations.

Field Programmable Gate Arrays Design, C programming using MATLAB for signal processing, Hardware descriptive language used: Verilog with C, , Xilinx ISE, Xilinx 8.1 and 13.3. Protocols worked: UVM, OVM, PCIe etc. Designing and verification of FPGAs. Designing Integrated Circuits using P-Spice