SUMMARY

With over 25 years experience in the full project life cycle of computer business systems, from meeting with users to handing over the system to maintenance developers. Industries I've been involved with include Real Estate, Finance, Healthcare, Government, and Insurance. Database expertise is in Oracle, both as a PL/SQL Team Leader, PL/SQL Developer, and Developer DBA. The sizes of these databases were as high as 5 TB. For Data Warehousing, we used ETL processes.

Skills Overview

• As a database developer / developer DBA, I have 16 years of DDL experience.
• Oracle 6 through 11G,R2 17 Years
• Creating and altering tables, indexes, sequences, procedures, packages, and triggers, etc.
• UTL FILE was used for data pump activities. On every project with 10g and higher, we either used data pump or imp, and exp. The server side capabilities and buffer bypass, along with u sing dbms job to call data pump jobs and avoiding creating sessions makes this tool great.

Sybase System 11.5, ASE 12.0, ASE 15.0

OTHER SKILLS

• Visual Basic 6.0, VBA, Com, Com, Dcom, Rational ClearCase, ASP, JSP, c, php
• C, C, PowerBuilder, Java, Clipper, PL/SQL, Netegrity Siteminder, Delphi
• C, Bourne, Bash, Perl, Korn, Sed, AWK.
• Solaris, Linux, HP, Windows 25 Years
• DB Architecture 3 Years
• CVSNT Administration 3 Years
• Erwin 10 Years
• PowerBuilder, up to v 7.0 9 Years
• Java 7 Years
• C, C 5 Years
• PHP 5 and 6, MySQL 5 3 Years
• Javascript 5 Years
• CSS 3 Years
• WAMP LAMP 2 Years
• AJAX 1 Year
• JSON 1 Year
• JQuery 1
• HTML/XHTML 5 Years
• JSON 3 Months
• Perl, OO Perl 3 Years
• ETL Pl/SQL 17 Years
• Oracle ODI 6 Months
• FOXPRO current 2.6 6 Years
• Allclear Flowcharter 1 Year
• Oracle 9i Forms 1 Year
• SCT Banner 1 Year
• Visio 5 Years
• Oracle OTL 2 Years
• Excel VBA 2 Years
• Data Warehousing Oracle 10 Years
• Operating Systems
• Shell Script Languages 15 - 20 Years
• Other Languages
• XML Development
• XSLT, XML, XSD, DTD, XML-Schema, Oracle XDK, XKU, XSD style sheets
• OTHER DATABASE SQL SERVER 5 YRS
• Paradox, Clipper, Foxpro, Dbase III, Dbase IV, Dataease, MS Access, Sybase, MS Sql Server, mysql.

Work Summary

Oracle Developer

Confidential

• Cross Session pl/sql function result cache for sharing the use of functions in memory in between sessions.
• Pl/sql hierarchical profiler. After the DBA setup the grants and created the test directories, profiling was easy.
• Dbms prof.hstart profiling, called some functions, and dbms hprof.stop profiling, dbms.analyze. After getting the runid from dbmshp runs, we used the table dbmshp function info to examine info for that runid.
• PL/Scope was used by setting the plscope settings parameter to identifiers:all to enable the collection of all used identifiers by selecting the user identifiers table with start and connect to gather information in a convenient order
• Used Name and Mixed Notation to better call functions and procedures with important parameters first with names, and unimportant parameters as defaults
• Used the Continue statement to better escape out of loops
• Used Automatic inlining by setting plsql optimize level parameter to 3 to have the optimizer automatically used inlining when necessary along with the inline pragma
• Used automatic native compilation by plsql code type to Native instead of the default, Interpreted. This does improve our procedure code.

Oracle 10gR2 New Features used:

• Conditional compilation was used by altering our sessions and setting plsql ccflags equal to debug:1. An example is if debug 1 then do something end.

Oracle 10gR1 New Features used:

• Pl/sql code was compiled for native execution, which translates code in to c code, and then compiled using a c compiler, and linked into the Oracle process, mostly at the session level using alter session set plsql code type 'NATIVE'.
• Used BULK ROWCOUNT with forall. Also used select with bulk collect and colletion objects.

Confidential

PL/SQL Developer, VBA

• At Estee Lauder, we have 35 brands of products in 170 countries, each combination is a schema. For each schema, for example, Clinique-China, we would call the schema clcnweb. These schemas would actually have 4 instances on 4 different servers, one for Development, Testing, Staging, and Production. The database had to be architected this way for CVSNT promotions. Many schemas unfortunately did not have all these for instances, this had to be fixed.
• At Estee Lauder, we have a several Terabyte warehouse. This warehouse stores all aggregated information for all 35 plus brands, for all 170 countries. On a daily basis, we have materialized views that are executed, which populate these historical tables for reporting. We use numerous bitmap indexes, partitions, pruning, and analytic functions to pull out this information, mostly sum over, ranking, etc.
• Our migration work is all based on fully documented ETL procedures. The first step is to meet with users, and build a mapping document between the old system and the new system. Before we load any data, we go through a data cleansing process, and load data into temporary tables. All data brought over is also logged into audit tables. We build ETL scripts and test for accuracy of data and that the process can be reliably repeatable. Another part of the process is making sure that the data is loaded as efficiently as possible. Because we are using PL/SQL for the ETL process, the process is very flexible and changeable. All processes take advance of parallelism, forall and bulk collect. Given that we often bring over numerous tables from an old system, we create migration scripts for each table, making the whole process very modular. Part of the data cleansing process is normalizing the data when appropriate.
• I spent several months working with Oracle ODI to help in working towards to a better approach. I created several source and target data stores. Setting up a meta repository. I also helped in setting up the fact and dimension tables. I created many bitmap indexes, partitions, and pruning.
• I created several spreadsheets using VBA for financial reports for management. These reports connected to their Oracle warehouse databases for data.
• Once the CVSNT process was fully architected, we had to then perform a data scrub of the whole organization, many instances of many brand-region combinations had source code that was different from development, so a large scale business activity ensued.
• CVSNT was incorporated into SQL Navigator via Team Coding connections. A build make shell script was created such that when any committed source code change would automatically be called to keep database source in sync. On promotions, this build make.sh script would take the changed source code and promote this source to the target environment. This was a very large architectural undertaking. This build make.sh script was able to determine the source db info and target db info, and bring the source code from the repository to the target environment. The CVSNT processes are running on a Linux VM and the database is Oracle 11gR2.
• Flow charts created to describe the whole CVSNT process using Allclear Flowcharter software. Earlier in the process of using CVSNT, Visio was used, we switched because management liked Allclear better.
• All pieces of this Oracle/CVSNT architected solution has been put into complete documentation. Documents for Administrators, promoters, and developers to perform their jobs using SQL Navigator tied to CVSNT.
• At Estee, we update each brand/countries website every 5 years, so there is always ETL migration work. Moving the older data to the newer systems and altering the data to fit the new system. I've done dozens of these migrations as well. Our migrations involve updating, adding or removing procedures, packages, triggers, functions, tables, indexes, views, etc. Our migration/ETL work is done using PL/SQL scripts. All reporting that we do takes advantage of reference cursors. All product catalog information for websites all over the world, are updated via Materialized views. When customers opt in to any of our websites, they have to accept certain criteria such as our privacy policy we use BLOB columns to store this information amongst over uses as well. To extract any of our privacy policy information, we use advanced XML. These migrations were 50 Oracle to Oracle Migrations and 50 Sybase to Oracle Migrations.
• One part of the tuning process is having a very good knowledge our data, as well as having a properly modeled data model allows us to take advantage of the best sql tuning strategies. All our websites take advance of Oracle RAC, we use 2 node clusters. When any product prices change, we use advanced queuing and Streams for storing and retrieving messages to know how to adjust prices for the particular brand and region.
• Many of our management sales reports are delivered using Excel and VBA.

Confidential

Oracle 11g PL/SQL development

This was a temporary assignment. Most time was spent performance tuning various running SQL statements. I fine tuned various packages for performance.

Confidential

Oracle PL/SQL development

• I worked on several projects including supporting many websites for database support. I created and altered numerous packages, procedures, triggers for these sites. Many management reports were also created.
• Jefferies and company has a huge, multiple Terabyte database which gets populated on a daily basis, mostly stock and bond information. Materialized views exist for populating these historical data. There are at least 6 dimensions minimally including counterparties, Fitch, Moody's and other ratings agencies. I did create reports out of this warehouse. I created many spreadsheets using VBA for product sales.
• I created several ETL PL/SQL scripts for data migrations. We brought in SQL Server and Sybase data into Oracle. This was in support of a large data warehouse. I performed data cleansing strategies for normalization, and business rules. I met with several business analysts.

Confidential

• I supported and enhanced a Mortgage back securities program written in Powerbuilder. I also created and supported, and tuned, several PL/SQL programs in support of this Mortgage program. I also created numerous Crystal Reports for this Mortgage program.

Confidential

Oracle DBA, Data Warehousing support

• This was a very temporary role as an apprentice to learn how to become a production Oracle DBA. I worked with Oracle database control and Oracle grid control. I also helped set up a program which could store all SQL performance information so that the firm would know where all bottlenecks were. We had automatic pulling of AWR information from Oracle into this repository for reporting purposes. The tool used was dbTrends. I helped in setting up ASM, and RAC configurations on Linux. I provided several migration / ETL scripts for bringing Oracle 10g data into Oracle 11g for several website upgrades. I was responsible for helping to populate our data warehouse by creating numerous materialized views. Data is populated twice a day via dbms mview procedures. We used bitmap indexes, partitions, and pruning.