MURRAY PUNG
Melbourne,VIC
Summary
Murray has advanced statistics qualifications and extensive programming experience. Notable strengths include understanding business problems and helping managers overcome them with advanced data analysis and visualisations. Murray is an expert in AB Testing, including sample size calculations, experiment design, and analysis
Overview
19
19
years of professional experience
Work History
Advanced Analytics Lead
Local Agent Finder
09.2023 - 05.2024
- Creating R Shiny App dashboards to support marketing, product, and account management teams
- Automated AB Testing modular framework implemented for product and marketing activities
- Writing of the analytical section of the monthly Board meeting
- This role has touched on all aspects of the business, requiring a deep understanding of the customer journey and the application of advanced analytical methods to identify opportunities for growth and business development
- Time Series Analysis (including SARIMA) and Generalized Linear Models were extensively used in this role
- Examples of projects include testing time series correlation between internal data and market data, also using Logistic Regression to identify significant factors that affect conversion rate (this work identified which house price range is most likely to result in a conversion
- Harmonising various data sources is essential to this role – external Core Logic data was used to complement internal results, and to further identify geographical areas of opportunity
- Technologies used include SQL, R, Python, Shiny Apps and MS Office.
Fraud Detection Data Analyst
NAB
12.2022 - 06.2023
- Provided analytical and data visualization support for a new Buy-Now, Pay-Later product.
- Developed relationships with key stakeholders, including non-technical product managers, to understand product challenges.
- Created a workflow to harmonize data from multiple sources using Splunk, Oracle, and Python, and developed a self-service report in Power BI covering product growth, delinquency, customer churn, and financial crime analyses.
- Conducted ad hoc analyses to identify significant factors associated with customer churn using A/B testing and advanced statistical methods.
Performed statistical analyses for fraud detection, including social network analysis, risk factor detection for financial product applications, and machine learning for fraud prediction. - Utilization of machine learning techniques for fraud detection, including BioCatch, Delta, and NetReveal.
- Developed fraud detection capabilities for a Buy-Now, Pay-Later product, including a dashboard for monitoring detection performance and analyses that uncovered multiple identity takeover syndicates.
Predicted delinquent accounts by selecting applicants with prior high risk of fraud behavior, which were subsequently confirmed as identity takeover fraud. - Forwarded identified cases to fraud investigators, who confirmed a high number of identity takeover incidents.
Conducted an internal fraud investigation involving the development of a Poisson Regression model to detect bankers with significantly higher fraud alerts. - This work confirmed systematic fraud (mostly fraudulent pay slips) committed by bankers.
Data Analyst
ANZ
06.2021 - 12.2022
- Using SQL, Python, PySpark, and Apache Spark, to redesign business processes, converting unstable, manually intensive programs to fully automated processes requiring no technical support.
- Analysis of customer lending behavior to predict default and refinancing
- SQL, R, Python, and PySpark were used to derive an analysis dataset from various sources.
- AB Testing, Random Forest, K-Nearest Neighbors, and Logistic Regression algorithms were tested to derive a model to predict the propensity to churn.
- A Cox Proportional Hazards model was also used to test the success of new products in reducing customer churn.
- DataCamp training, focusing on explanatory analysis and prediction methods
- Function writing in Python and R
Data Analyst
Novotech
03.2020 - 06.2021
- SAS, SQL, and R programming for extraction, transformation, and analysis of raw data
- Leader of multiple simultaneous projects, and provide support to colleagues as needed
- Created a monthly support meet-up named 'Programmer Happiness' that aims to share technical and business process knowledge
- Creation of dashboards for internal and external clients
- Transformation of clinical study data
- Use of standardised programs to perform repetitive tasks
- Analysis of business process to identify potential improvements.
Product Data Analyst
99designs.com
10.2018 - 03.2020
- Using SQL, Python, R, and AWS Redshift, this role provided analytical support to product managers across the organisation
- Broadly speaking, I worked with product managers to understand their business problems and help them solve them using best practice data analysis
- An example of an analysis that resulted in reduced designer and customer churn includes the discovery that the distribution of invitations from customers to designers was heavily skewed
- A relatively small number of designers were receiving the bulk of invitations and could accept only a small proportion of invites
- This resulted in customer churn due to many unanswered invites (low unanswered invites was highly correlated with customer churn), and inversely, low invite counts was correlated with churn in top designers
- The analysis identified the problem, but I also used analytics to help reduce the problem - when customers searched for designers, after certain filters were applied, generally the same designers were presented to customers
- This was the general cause of the skewed invites distribution
- I proposed a redistribution and tested two new designer recommendation engines (both applied after the customers' filters were applied)
- One gave weight to designers with lower invite count using a Decision Tree machine learning algorithm, the other was completely randomised
- I then proposed an AB test which found both recommendation engines to reduce churn in designers and customers
- Other examples include design and analysis of A/B Tests to test new product features, segmentation analysis and time to event analysis (designers were segmented into platinum, gold, silver, and bronze
- The time to churn was found to be sooner in the platinum designer segment, thus a designer retention strategy was developed specifically for platinum designers), pricing analysis to assess new price structures, detection of positive and negative associated factors, and improved fraud detection
- I was solely responsible for deriving analysis datasets, selecting the appropriate analysis method and presenting results to a non-technical audience.
Product Data Analyst
Xero
10.2016 - 10.2018
- Using SQL, R, and Google Big Query, this role provided data analysis support across the organisation, including product data analysis and money laundering detection
- Often this involved regular reporting of metrics, A/B Testing (including experiment design and analysis) to assess new product features and statistical analyses to help product managers better understand customer behaviour and to identify areas of potential improvement
- Benford's Law, and outlier detection were used to detect money laundering
- Segmentation analysis and Logistic Regression was used to create customer cohorts based on product usage (daily usage, tax time only, etc)
- The segmentation analysis was used to create cohort specific customer onboarding
- Logistic Regression found the tax-time only customers were converted at a higher rate
- I primarily used R, SQL, and AWS Redshift to perform data manipulations, analyses, and to present to stakeholders
- Stats techniques I've applied in this role include Survival Analysis to detect factors leading to customer churn, Fisher's Exact Test to test conversion rates between existing and new product features, Logistic Regression and other Generalised Linear Models to test product upgrade impact on conversion
- I mentored two junior analysts and am an active member of Xero's analytics community
- The data sources of this role include relational databases, Google Analytics, and BigQuery
- This requires data programming of large datasets with significant efficiency requirements
- I was solely responsible for applying appropriate statistical analyses to support stakeholders' business objectives.
Consultant Data Analyst
Servian
03.2015 - 10.2016
- This consultant role has developed my skills to include machine learning techniques within a big data environment, UNIX Shell scripting, and general communication
- I regularly use R, SQL, and SAS to perform complex analyses
- Details of my deployments are below
- SAS Server migration for a major Telco which involved UNIX shell scripting, SAS Base Programming, SAS Data Integration Studio, SAS Management Console, and SAS Enterprise Guide
- The project involved liaising with end users and project managers to ensure smooth migration of SAS objects and the implementation of a new file structure
- This involved extensive documentation and communication
- ETL with a big four bank's Enterprise Data Warehouse, and reporting using SSRS
- Marketing Analytics role at a major loyalty program
- This included SQL on a transaction database in support of marketing campaigns
- Oracle, SAS, and SQL are used to derive campaign lists and analyse their efficacy
- RFM model used to optimise customer targeting
- A marketing analyst role for an Internet Service Provider
- Tasks include programming and analysis for marketing lists, post campaign analysis, churn, and business strategy (particularly NBN rollout)
- Tools used include R, Adobe Analytics, SAS, Python, UNIX.
Data Analyst
Actelion
10.2012 - 02.2015
- Validation of CDISC SDTM data conversion
- This role was focused on data quality for FDA submission
- I wrote SAS programs to convert raw datasets to SDTM format
- Common techniques include PROC SQL & the datastep, arrays, macro variables, date conversion, transpose, PROC REPORT
- The role was highly technical but effective communication with colleagues was an essential responsibility.
Consultant Data Analyst
01.2011 - 10.2012
- I worked as a Statistical Consultant within the insurance and banking industries
- Techniques applied include Credit Risk Scoring and other risk analysis techniques
- I also redesigned a number of business reporting workflows to better serve stakeholders
- One project included transforming a manually intensive reporting process into an automated system using SAS
- Roles included credit risk analysis for the unsecured loans teams at NAB, statistical analysis of workplace injury data at WorkSafe Vic, and reporting transformation at AAMI / Suncorp.
Statistician and SAS Programmer
CSL Limited
06.2009 - 12.2010
- Contract role as a Statistician and SAS programmer
- My role involves data analysis for internal and external clients, managing and querying data, advanced statistical analyses, creating tables, figures and listings for FDA submission and ICH reports, ad hoc SQL queries, statistical advice and experiment design, standardised SAS program and macro development
- SAS and R Programming, for statistical analysis (ANOVA, Logistic Regression & various Generalized Linear Models, Non Parametric tests, diagnostic checks of data models)
- Data transformation to standardised datasets
- Writing statistical analysis reports for a non technical audience
- Extensive SQL programming for data extraction and transformation
- Independently working with internal customers to provide data analysis relevant to their business needs
- Developing a standardised format for SAS programs to increase understanding between programmers, increase efficiency and provide auditing capability
- Providing Help Desk support for other SAS and R users within CSL Limited
- Using R for statistical analysis and visualisations
- Coordinating projects with other statisticians.
Data Analyst
GlaxoSmithKline
01.2008 - 06.2009
- Create standardised SAS programs using standard and study specific macros to produce Immunogenicity, Safety/Reactogenicity, Demographic, CTRS, and individual data listings for the Statistical Report
- Writing Statistical Analysis Reports
- Manage numerous projects simultaneously in a highly regulated environment
- Implementation of Clinical Data Interchange Standards Consortium's (CDISC) Study Data Tabulation Model (SDTM) for the Trial Design Data and Domain Models using SAS DI Studio
- SQL and SAS programming for data extraction and analyses
- Quality control of team outputs
- Produce statistical reports for a non technical audience.
Data Analyst
Datapharm Australia
01.2005 - 01.2008
- SQL, SAS, and R programming for data extraction and clinical trial analysis
- A/B testing to assess the efficacy and safety of clinical trials
- Advanced statistical methods such as Generalised Linear Models and Survival Analysis (to assess if treatments were effective).
Education
B. Information Science (Statistics) -
University of Newcastle
Newcastle, NSW, Australia12.2004
Grad Cert Biostatistics -
University of Sydney
Sydney, NSW, AustraliaData Camp (Data analysis, machine learning, Shiny App development) -
Skills
- Statistical data modelling
- AB Testing and Experiment Design
- SQL
- Python
- R
- SAS and JMP
- Stata
- Excel
- Data visualisation [Shiny, GGPLOT]
- Acquisition & churn analysis
- Customer segmentation [and other machine learning methods]
- Cloud platforms [Google Cloud Services (Analytics and BigQuery), Snowflake, AWS Redshift, MS Azure, Amazon S3, Spark]
Additional Information
15+ years of advanced data analysis experience
Personal Information
Title: SENIOR DATA ANALYSIS SPECIALIST
Timeline
Advanced Analytics Lead
Local Agent Finder
09.2023 - 05.2024
Fraud Detection Data Analyst
NAB
12.2022 - 06.2023
Data Analyst
ANZ
06.2021 - 12.2022
Data Analyst
Novotech
03.2020 - 06.2021
Product Data Analyst
99designs.com
10.2018 - 03.2020
Product Data Analyst
Xero
10.2016 - 10.2018
Consultant Data Analyst
Servian
03.2015 - 10.2016
Data Analyst
Actelion
10.2012 - 02.2015
Consultant Data Analyst
01.2011 - 10.2012
Statistician and SAS Programmer
CSL Limited
06.2009 - 12.2010
Data Analyst
GlaxoSmithKline
01.2008 - 06.2009
Data Analyst
Datapharm Australia
01.2005 - 01.2008
B. Information Science (Statistics) -
University of Newcastle
Grad Cert Biostatistics -
University of Sydney
Data Camp (Data analysis, machine learning, Shiny App development) -
Similar Profiles
Bronwyn LinnaneBronwyn Linnane
Client Service Consultant at Local Agent FinderClient Service Consultant at Local Agent Finder
Bronwyn LinnaneBronwyn Linnane
Client Service Consultant at Local Agent FInderClient Service Consultant at Local Agent FInder
Alexander HowellAlexander Howell
Realtor at Local Agent (formerly Chief Properties)Realtor at Local Agent (formerly Chief Properties)