SUMMARY

• Skillful Data Scientist with over 10 years of experience in Information Technology / Python / R/Looker/Tableau and dedicated 4.5 years of experience into data science, machine learning, natural language processing (NLP), & predictive modelling.
• Proficient in structured / unstructured data, data modeling, data mining, and data profiling.
• Successfully completed 5 data science projects
• Expert in data cleaning, features scaling, features engineering using Pandas and NumPy packages in python
• Extensive experience in developing different statistical machine learning, data mining solutions to various business problems and generated data visualizations using R / Python
• Proficient in machine learning algorithms and statistical modeling like decision trees, text analytics, natural language processing (NLP), supervised / unsupervised, regression models, social network analysis, neural networks, deep learning, SVM, and clustering using scikit - learn package in Python / R
• Well-versed in Predictive Modeling with SAS, R, and Python
• Expertise in MLlib, Spark's Machine learning library to build and evaluate different models
• Automated Data Import Script using shell scripting, PHP, MySQL and regular expressions.
• Skillful in Natural Language Processing (NLP) for speech recognition and used word to vec to understand the association of the words
• Worked on Jupyter notebook to visualize stores using maps

TECHNICAL SKILLS

• Data Science
• Python
• R
• MySQL
• Web analytics
• Clustering & Segmentation
• SQL
• Python
• RStudio
• PostgreSQL
• DataBricks
• PySpark
• GIS
• PostGIS
• Flask
• Databricks
• Integration Services (SSIS)
• Oracle 9i OLAP
• MS Office Web Components (OWC11)
• JDBC
• HTML5
• DHTML
• XML
• CSS3
• Web Services
• WSDL
• Erwin R 9.6 9.5 9.1 8. x
• Rational Rose
• MS Visio
• Spark peg
• Hive
• HDFS
• Map Reduce
• Pig
• Kafka
• SQL
• Hive
• Impala
• Pig
• Spark SQL
• SQLServer
• MySQL
• MS Access
• HDFS
• HBase
• Teradata
• Netezza
• MongoDB
• Cassandra
• SAP HANA
• MS Office (Word/Excel/Power Point/Visio)
• Tableau
• Rstudio Markdown
• Business Intelligence
• SSRS
• SVM
• GitHub
• Tableau
• Azure Data Warehouse
• Windows
• Linux
• Unix

PROFESSIONAL EXPERIENCE

Confidential

Data Analyst

Responsibilities:

• Analyzed data to tell astory (what happened, why it happened, what caused the numbers) and Recommended (what should the leadership do to improve)
• Performed data analysis to support and implement business processes and decisions
• Perform analytical deep dives to analyze problems and opportunities, identify the hypothesis and design & execute experiments and provide recommendations into key performance metrics and goals
• Designed, created, and maintained dashboards for KPI monitoring and identifying trends & discrepancies
• Developed a deep understanding of customer journey phases and key business metrics. Understand how and why customers engage with your product.
• Enhanced UX by using segment debugger and analyzing the API calls
• Utilized RedShift, SQL, Looker and tableau tool on a daily basis to provide analytics support across the business, including sales, product, marketing and executive teams
• Created efficient PDTs and improve run time of the looker dashboards
• Evaluated success of product initiatives by setting success criteria and analyzing results
• Provided analytics perspective to aid in product planning
• Good understanding of data architecture process
• Troubleshoot within backend developer tool to resolve BI issues
• Defined milestones, deliverables and communicated project scope, requirements
• Acted as liaison between clients and technical teams
• Developed communication plan and improved team collaboration between onsite and offshore teams
• Worked on ad hoc requests from internal clients.
• Documented functional and business requirements in collaboration with the product development team
• Worked with project or functional leads in developing use cases/ scenarios, flow diagrams and performing workflow analysis
• Managed cross-functional projects by providing project leadership and daily management throughout the project from inception to delivery
• Managed the feature development and defect backlog on a daily basis to ensure that priorities adhere to the strategic direction outlined by the Product Manager
• Work closely with Product Management group to provide data solutions to issues impacting customers
• Assisted with building training material content for new functionality developed for Collections Management
• Translated complex concepts into implications for the business via excellent communication skills, both verbal and written
• Extract meaningful insights through analyzing large, complex, multi-dimensional customer behavior data sets
• Identify key trends and build executive-facing dashboards to track the progress of acquisition, monetization, and engagement trends.
• Inform future experimentation design and roadmaps by performing exploratory analysis to understand user engagement behavior and derive insights

Confidential, Bradenton, FL

Data Scientist

Responsibilities:

• Performed data cleaning and feature selection using MLlib package in PySpark on Databricks
• Performed data cleaning, features scaling, features engineering using pandas and NumPy packages in Python
• Used NLP for sentiment analysis for the insurance and medical records
• Worked on Jupyter notebook to visualize stores using maps
• Built Factor Analysis and Cluster Analysis models using Python SciPy to classify customers into different target groups
• Used clustering techniques like DBSCAN, K-means, and Hierarchical clustering for customer profiling to design discount plans according to their behavior pattern in R and Python
• Performed spatial analysis using QGIS and manipulating the geometry types to visualize the data using shape files
• Used GIS (geographic information system) framework for gathering, managing, and analyzing data
• Generated visual maps using map projections in QGIS and used Confidential analytics
• Performed data preparation, loading, and management in an QGIS PostgreSQL server environment
• Developed good understanding of SRID, CRS, SRS, Vector Geometry, and Geodata (GeoJSON / WKB / WKT formats)
• Automated manual MS Excel tasks using macros which helped boost productivity
• Used VLOOKUP on raw excel file and other solver tools for better efficiency
• Created / designed reports that will use gathered metrics to infer & draw logical conclusions of past and future behavior
• Used containers like Docker for version control and seamless accessibility for clients
• Built Docker image to preprocess data and deploy model as an API
• Used R packages like BSTS (Bayesian structural time series), Boom, BoomSpikeSlab to infer Causal Impact and compare control group vs test group in a post period intervention in RStudio
• Generated graphs and reports using ggplot package in RStudio for analytical models
• Used RStudio Markdown for reporting and storytelling
• Worked on application of various machine learning algorithms and statistical modeling like decision trees, text analytics, natural language processing (NLP), supervised / unsupervised, regression models, social network analysis, neural networks, deep learning, SVM, clustering to identify volume using scikit-learn package in Python / R
• Used Principal Component Analysis in feature engineering to analyze high dimensional data
• Used NMF algorithm to produce latent features of an ACS dataset and plot it on a map to understand similarities between each state to further use it in optimized targeting
• Used Agile methodology and Scrum process for project developing

Confidential

Data Scientist - I

Responsibilities:

• Worked in a team of Data Scientists/ML Engineers to build and deliver multiple machine learning applications & data products with most of prototypes done in Python
• Developed multiple recommender systems that helps customer build customized vehicles, recommend warranty packages, parts, services, and used Clickstream data analysis
• Used Natural Language Processing (NLP) for speech recognition and used word to vec to understand the association of the words
• Used Jupyter notebook for Python coding
• Defined end-user experience and benefits individual recommender system
• Evangelized needs and benefits of recommender system to executive managements
• Presented business insights about the user behavior and product behavior
• Used macros for raw excel data files and made it useable for analytics
• Worked with team to develop state of the art recommender system using customer history /product history and used Kubernetes for scaling & development
• Evaluated multiple models based on affinity analysis, Deep & Wide Neural Network, ALS Matrix Factorization Model, Hybrid Collaborative Filtering with user based, and item-based models
• Formulated A/B testing metric and designed automated scheduled workflow to perform continuous A/B testing
• Deployed recommender models through web APIs to allow downstream applications for each new version of recommender system to easily consume recommendations
• Reduced custom built configuration time by 30% using the model
• Increased product (parts, warranty package) sales by 12% using the model
• Developed & deployed AI model that performed NLU (natural language understanding) on customer complaint and used vehicle configuration to predict repair hours, failed component/part, and rank procedures
• Performed data manipulation, data visualization, and feature engineering
• Performed feature engineering on text using word2vec model andhandled imbalanced class problem
• Built multiple ML models like Fully Connected Multi-Layer Neural Networks, CNN, ANN, RNN-GRU, Memory Network, Conv1D, and GLM
• Used libraries like TensorFlow and PyTorch(Rtorch with R)
• Evaluated model on multiple evaluation metrics like Top-K Mean Average Precision, KAPPA score, Precision,and Recall
• Predicted from more than 4.8K unique vehicle components/parts using model and achieved 82% accuracy on test set
• Deployed the model successfully on multiple downstream applications used by customer, dealership, technicians, and fleet managers
• Reduced average vehicle issue diagnosis time by 40% using model

Confidential

Junior Data Scientist

Responsibilities:

• Built Dynamic Forecasting Models across multiple dimensions served in real-time & batch process e.g. Weekly Vehicle Retail Forecast at models, categories, region & dealers, monthly/weekly warranty &parts forecast
• Identified the use case and laid out quantifiable benefits & clearly defined risks
• Developed / implemented R and Shiny application which showcases machine learning for business forecasting
• Performed validation on machine learning output from R
• Coordinated with Senior Director and Executives to present analytics workflow
• Co-led the team to identify machine learning approaches that could deliver quick results
• Evaluated merits of individual ML approaches and established evaluation metrics
• Built forecasting models like Forecast Hybrid, STL Arimax, RNN-LSTM,RNN-GRU, hierarchical &grouped time series, dynamic regression, vector autoregressions
• Formulated end-to-end solution and monitored team’s progress
• Delivered successful forecasting model with 85% accuracy for 12 month’s forecast horizons

Confidential

Ad-Ops Engineer

Responsibilities:

• Performed trafficking and delivery of all advertising related projects
• Managed digital advertising campaigns including Ad trafficking, packing, optimization, reporting, and monitor campaign progress
• Handled various tracking problems, ad tag implementation effectively
• Developed solutions / documented internal procedures, policies, and tutorials regarding ad tracking & targeting for DoubleClick, AdMob, AdSense / AdWords
• Worked with Product Management, Marketing, Engineering, and other teams to inform, test, and implement product and operational improvements
• Optimized campaign performance by recommending adjustments to inventory and creative components of the campaign

Confidential

Data Analyst

Responsibilities:

• Worked with data governance, data quality, data lineage, and data architect to design various models.
• Delivered MS Excel tasks using macros
• Analyzed large and complex datasets to investigate and identify fraud scheme trends and improve efficiency in fraud detection process.
• Classified fraud accounts to deconstruct emerging patterns.
• Built and managed reports and dashboards to monitor KPIs.
• Prepared technical design documents and test cases