University of Sydney

Baulkham Hills High School

Data Scientist at Ensu

- As a data scientist I am responsible for working with clients on a variety of B2B projects leveraging our unique transactional data set, maintaining and augmenting the data asset and working with internal teams to help inform product and marketing/growth decision making.

Pogo

Data Scientist

Programmatic Merchandizing (July 2022 - 

-As the PM/data scientist I worked to bring this new vertical to market going from 0 -1 and growing the business from 0 to 15% of all travel bookings in the first year.  
-Managing + implementing personalized travel recommender system leveraging AWS Personalize - responsible for generating $3 million in revenue for the business.
-Designed, launched + owned 4 new features, generating more than $20 million revenue. 

Rental Cars (March 2021 - June 2022)

-Built + deployed end-to-end models to predict probability of cancellations + no-shows based on user + booking data, providing a view of net revenue at time of booking.
-Managed + implemented current notification strategy, driving liquidity (+20%) and top line bookings (+15%), contributing to over $3.5M in revenue
-Designed, implemented + managed a new “Price Match” feature from inception to MVP.
-Developed new “Price Drop” notifications - highest converting notification accounting for 5% of bookings + 780k revenue.
-Built data collection pipeline + dashboards/reports to monitor + measure pricing competitiveness.
-Managed, designed + implemented ETL process, including framework to standardize + ingest supplier data into warehouse.

Hopper

Senior Data Scientist

-Responsible for designing, implementing and managing entire data science pipeline from Survey Design & Data Collection to ETL, Modeling & Deployment 
-Built deep learning ensemble consisting of autoencoder + classifier using Tensorflow, iteratively  achieving more than 300% uplift in accuracy over initial rules based prediction 
-Built and implement ensemble recommender system consisting of collaborative filtering/matrix  factorization + autoencoder combined with clustering 
-Worked closely with Product to create metrics for retention, flow and funnel analysis and automating  this through Mixpanel integration with BigQuery and R 
-Set up backend framework to pull data from APIs and load into database using Celery and Django

Ensu

-Conducted comprehensive churn analysis of customer base using combination of machine learning (XGBoost, Random Forest) and statistical (Lasso, Logistic Regression) models. Identified leading ‘triggers’ that are highly indicative of churn and could potentially save more than $100 million
-Used customer address information to geo-code addresses and combined this with census data to add 100+ geo-demographic features for use in modelling.

Rest

Analyst - Data Science

-Worked with clients across a variety of industries on ad-hoc data science consulting projects.
-Researched and implemented latest machine learning techniques in R and Python for analysis.
-Developed statistical models (Linear/Logistic regression, Additive Models, Mixed & Random Effects) and machine learning models (GBM, XGBoost, Deep Learning) using R and Spark to model  a variety of classification and regression problems in risk and marketing.
-Forecasted future credit demand using time series models (ARIMA/ARIMAX).
-Conducted complex statistical analysis using internal big data platform (Hadoop & Spark).
-Used latest visualisation tools (Tableau, ggplot2) to deliver complex insights to clients.

Equifax

Analyst - Client Analytics (Data Science)

Obtained, cleaned and analysed world economic data and assisted in modelling using R & Stata.

Nishant Patra