SUMMARY:

• Data scientist with deep understanding, knowledge and experience in Machine Learning, Deep Learning, Neural Network, data analysis, trend analysis and predictive analysis .
• Big data/Database expert with proven success in designing and developing complex and large scale database/Big Data system.
• Business - minded data scientist with a demonstrated ability to deliver valuable insights via data analytics and advanced data-driven methods.
• 20 years of hands-on experience in high-performance data analytics, reporting, modeling, performance tuning and on databases/data warehouse/Big Data System.
• High familiarity and sound understanding of big data/database methodologies, tools, and development life cycle (SDLC) programming standards and industry best practices.
• Excellent analytical skills and problem-solving capabilities in system integration that involves heterogeneous technologies.
• Extensive experience in government consulting.

TECHNOLOGIES AND TOOLS:

• Supervised Learning, Unsupervised Learning, Reinforcement Learning.
• Classification, regression
• Forward propagation, backward propagation.
• Gradient Descent
• Activation functions
• Model optimization.
• Decision trees: Random Forest, Gradient Boosted Trees.
• Python, NumPy, Pandas. Scikit-learn, MatPlotLIb, Jupyter Notebook
• TensorFlow,
• ScikitLearn.
• Bias, Weight, Activation.
• Recurrent NN, Convolutional NN, LSTM. NLP
• Dropout.
• Hyper parameters.
• SparkML, SparkDL, SparkSQL, Structured Steaming.
• Data frames. RDD. DAG. Lazy Evaluation
• HDFS. Map/Reduce
• Hive, HQL
• SAS
• Clouderea
• Impala
• Pig
• Data mining and modeling
• Predictive analysis
• Trend analysis
• Data visualization
• SAS
• Oracle 8/8i/9i/10g/11g/12c.
• Database design, ETL, reporting, performance tuning, data analysis, Database development, business intelligences, data integration, configuration, architect, modeling, DBA.
• PL/SQL, SQL, UNIX Shell Scripting, XML, JSON, Java
• My SQL,SQL Server. data warehouse
• NoSQL database, MongoDB.
• Linux, UNIX, MS Windows, DOS.
• Chef, Ruby, Vagrant.
• Liquibase,Jenkins, GIT, SVN.
• TOAD, Data Pump.
• Oracle Enterprise Manager, Oracle Designer, Oracle SQL Developer Data Modeler, OBIEE, ERWin, Oracle Warehouse Builder, Oracle Analytic Workspace Manager, X Manager, Informatica, SQL*Plus, SQL Navigator, Oracle SQL Developer, Oracle Discoverer, Oracle Express, Oracle SQL Developer.
• Agile Methodology

WORK EXPERIENCE:

Senior Data Scientist

Confidential

Responsibilities:

• Implemented data mining and enterprise resolution.
• Implemented data visualization with MatPlotLib and Pandas.
• Proposed Strategy/Road Map for Advanced Analytics:
• Data ingestion
• Machine Learning
• Streaming
• Spark
• NiFi
• Hadoop
• Data Lakes
• Developed processes for trend analysis, statistic analysis, time series and fuzzy matching.
• Furnish executive leadership team with insights, analytics, reports and recommendations, enabling effective strategic planning.
• Interact with client to gather requirements, get feedback and provide /brown bag.
• As part of the internal R&D project, did research and wrote white paper on resource efficient Neural Networks.
• Analyzed and processed complex data sets using advanced querying, visualization and analytics tools.
• Performed data analysis and data extraction.
• Developed machine learning models for predictions, using algorithms such as linear regression, Support Vector Machine, Random Forests, Gradient Decent, and Logistic Regression.
• Fixed data quality issue due to data format mismatch, source data quality, database design issue
• Developed ETL process using SAS
• Implemented work flow optimization on ETL processes that reduced the processing time from 42 hours to 10 hours.
• Did extensive research in Chinese Underground Banking.
• Performed text analysis.

Data Scientist

Confidential

Responsibilities:

• Implemented Big Data Analytics capabilities for trend analysis, streaming analysis, data integration, using Spark, HADOOP, Big Insight, Kafka, etc.
• Implemented data optimization of data ingestion, storage, computation, retrieval and provide a roadmap for future data analytics.
• Implemented Machine Learning for intrusion detection, using Spark ML.
• Applied SparkSQL, SparkML, decision tree ensembles, such as random forest, gradient boosted tree, vector assembler, Machine Learning pipeline, parameter grid builder, regression metrics evaluator etc. implementing Deep Learning on object recognition, using SparkDL
• SparkDL, pyspark.sql.functions, pyspark.ml, Pipeline, readImages, DeepImagePredictor,
• Tensorflow, TFImageTransformer,
• Keras, InceptionV3
• Numpy
• Converted relational database to MongoDB for data ingestion and processing.
• Implemented new functionalities for STIX 2.0.
• Implemented Oracle XML and JSON functionalities for CIAP project.

Data Scientist /Big Data Architect

Confidential

Responsibilities:

• Big Data Analytics with Hive, Pig, Impala, Spark, HDFS, Raw XML (Sequence Files) using Cloudera Framework
• Involved in Hive database architect, modeling and design on a large and complex system to generate reporting for a health care system.
• Developed work flow and ETL process
• Established standards/best practice for ETL patterns, batch patterns, exception handling and logging.
• Developed high-efficiency reports using very complex HQLs.
• Created ad hoc reports for clients.
• Did extensive performance tuning on ETL and reporting processes.
• Performed in-depth data analysis and profiling on different data sources that provides insights and facilitates reporting and trend analysis.
• Developed proof of concepts for new designs/tools.
• Was able to deliver under very pressing timeline with high quality work that repeatedly won very positive feedback from the clients.
• Linux Shell scripts for automating data loads from HDFS using Hive. Pig for XML sequence file analysis, used piggybank.jar and Python UDF’s for processing data. Created UDF’s in Hive for efficient data processing
• Worked to build a scalable platform to ingest and analyze large amounts of data with constantly changing analysis and reporting needs. Work with Hadoop HDFS and Hive provide highly scalable and robust Data Warehouse for data in virtually any format (Raw, Transformed/Processed, MPP capabilities for processing/transformation of data)
• Analyze/Validate data models against the XSD provided by technology partners. Automate loads using Shell scripting.
• Working with Petabytes of data, use data compression for efficient storage and I/O activities and Kerberos implementation for security layer.

Environment: Cloudera, Hive, Pig, Impala, Spark Linux(Shel/lPerl), SVN, XML& sequence files, Avro, Oracle

Data Architect / DBA

Confidential

Responsibilities:

• As a member of the architect team, established standards/best practice for ETL patterns, batch patterns, exception handling, logging, DML error logging, and deployment methods.
• Developed high-efficiency reports using very complex SQLs.
• Provide DBA support to 7 development, testing and architecture teams. DBA duties include: database installation, database configuration and integration, database user management, backup and recovery, set up restore point, flash back, data analysis using complex SQLs.
• Implemented performance tuning methods such as: indexing, query optimization, explain plan, query rewrite, parallelization, work flow optimization, bulk load, analytic functions etc.
• Worked very closely with DEV Ops team, supported builds and deployments, created deployment scripts and Liquibase scripts.
• Constant interaction with clients, did ad hoc queries and data analysis.
• Developed stored procedure for ETL and test data creation.
• Developed logical and physical model for TMIP database.
• Developed ETL in PL/SQL stored procedures.
• Redesigned database and data process to ensure data accuracy and improve performance.
• Installed and upgraded Oracle database, e.g. upgraded oracle 11.0.3 to Oracle 12.0.1. installed quarterly patches
• Set up multi-layered database security features. Performed extensive analysis to identify database configurations and settings that are not compliant with DoD standard. Fixed the issues by updating database configuration and parameter settings, and implement Oracle Advanced Security features.
• Developed efficient backup and recovery processes for the Oracle database.

Confidential

NOSQL Database Developer

Responsibilities:

• Installed,designed and developed MongoDB database for Confidential data store which integrates data from multiple sources. Applied performance tuning techniques such as indexing, sharding etc.
• Developed a process that uses Mongeez to deploy database changes using changesets in XML format.
• Extensive research in NoSQL database matrix with comparison of MongoDB, CouchDB and CouchBase.
• Developed cookbook for CouchDB installation. Installed and monitored CouchDB. Did view creation, data import and data analysis.
• Member of an architect team to define the approach for developing and implementing an architectural and implementation framework for the development of an Integrated Electronic Health Record application (iEHR). Main responsibility includes data standards, technologies and tools needed to establish the database part of the Confidential project.
• Developed reporting system called REXUS. Developed the logical and physical model of multi- dimensional REXUS database, using data warehouse star schema approach.
• Developed the ETL process that populates the dimensional model which enables quick and efficient reporting and performed DBA duties such as creating tablespace, creating users/schemas, creating roles, grants, managing tablespace, etc…
• Developed the ETL package, identified performance bottlenecks and did performance tuning on work flow and SQL queries.
• Developed data model for Enterprise Data Warehouse and developed PL/SQL package for ETL.
• Did extensive performance tuning, tested on various tuning methods, including data flow optimization using materialized view/ global temporary tables, incremental loading, table partition, partition exchange on large tables, parallelization of SQL and batches, indexing, bulk load, etc.

Principal Database Engineer

Confidential

Responsibilities:

• Performed full system development lifecycle (SDLC) tasks including MAILDAT database design, ETL, reporting, data analysis and ad-hoc request.
• Developed reporting packages that create report that provide internal and external customers with data on mail source, mail volume, mail location, postage statement, etc.
• Implemented performance tuning methods such as indexing, query rewrite, materialized view fast refresh and table partitioning as well as did performance tuning on 20+ reporting packages for custom reports.
• Designed an archiving process and developed stored procedures that archive historical data, which significantly reduced table sizes in the currently running system and improved performance.
• Completed table redefinition on a large number of tables in multiple schemas to convert non-partitioned tables to partitioned tables, and de-fragment some of highly fragmented tables and supported DBA teams, CM(change control) teams and 2 testing teams, in an effort to coordinate different team efforts to ensure smooth process.

Senior Database Developer

Confidential

Responsibilities:

• Main responsibilities included Oracle data mart design, ETL, performance tuning and unit testing. led a data mart team that designed data mart, and developed PL/SQL stored procedures for ETL and reporting on financial data for Global Business System project.
• Maintained excellent feedbacks in ETL high level/detail level design reviews and code reviews and demonstrated strong ability to meet short deadlines, deal with constantly changing requirements, coordinate with multiple teams and multi-task, as required in fast-paced projects.
• Worked on very high-profile, extremely fast-paced GFAS project with very pressing deadlines involving database design, ETL, performance tuning and CQ fixes for financial reports.
• Consistently received highly positive feedback from the client manager.

Enterprise Data Warehouse Architect

Confidential

Responsibilities:

• Performed full system development lifecycle (SDLC) tasks including data warehouse architecture, design, modeling, ETL, reporting, data analysis and ad-hoc request IMS(Inventory Management system).
• Did performance tuning, Improved the performance of ETL, reduced the processing time from about 20 hours to 6 hours by implementing: system re-architecture, incremental loading method, materialized view fast refresh, query rewrite, work flow optimization, bulk load, parallel processing and table partitioning (list partition and range partition).
• Integrated and merged 5 data warehouse applications into one mega mart and dramatically improved performance, reduced process time, eliminated conflicts and redundancy across systems and satisfy users from different systems with one application.
• As the Lead for the Operational Data Store (ODS) application, I was responsible for providing operational data to hundreds of users on a daily basis. I developed and maintained applications that comprises of 10s of thousands of lines of codes for ETL to load 10s of Millions of records as well as developed reporting system for users.