Jun 2022 – Present · 3 yrs 11 mos

Jan 2022 – Jun 2022 · 5 mos

Jan 2021 – Jan 2022 · 1 yr

Jun 2020 – Jan 2021 · 7 mos

Jan 2019 – May 2020 · 1 yr 4 mos · Amsterdam

• Building a recommendation system for Next best campaign and Next best personal offer to increase engagement
• Converting user journey to input features using pandas and SQL on AWS
• Identifying engagement models to generate supervised labels
• Using random forest for baseline recommendation results using aggregate historical data
• Using time series neural networks for recommendation improvement
• Identifying model success with stratified segmentations based on country, time , campaign type
• Working with IT team to track all digital foot prints of the end user under a systematic approach
• Working with product owners to deliver routine engagement metrics
• Collaborating with product owners to improve recommendation system and increase acceptance
• Hiring and Coaching team members

Jun 2018 – Jan 2019 · 7 mos · Amsterdam Area, Netherlands

• Created a ranking algorithm for Customer Segment of 1 from her customer journey for credit card fraud
• Used self supervision class of algorithms
• Trained with cross validation on 200,000 transactions with only .1% positive labels
• Developed a control parameter that balances the false positive and false negative rates
• Created a scoring system for big data vendor selection within oil and gas sector with approval of 15 architects from client
• Created 300,000 Euros client partnership for building analytics platform within insurance sector

Feb 2017 – Apr 2018 · 1 yr 2 mos · Amsterdam

• Built a Anomaly detection model using self-supervised learning
• Awarded 120,000 Euros in research from Dutch Innovation Committee to discover anomalous
• patterns by unmasking masked partial features and observing the representation in
• hierarchical clustering , using GRU based neural network
• Led a team of 5 to build various data science products such as geo location prediction , reporting ,
• content recommendation

Jun 2016 – Sep 2016 · 3 mos · Malta

• Built a random forest classifier from on 300 GB data to identify irresponsible gamblers 1 week in advance with .86 AUC to avoid negative press and its impact on stock price
• Built a random forest classifier and a regressor to predict life time value of a new gambler from 1 month of interactions

Feb 2016 – Jun 2016 · 4 mos · London, UK

• Built a random forest classifier from on 300 GB data to identify irresponsible gamblers 1 week in advance with .86 AUC to avoid negative press and its impact on stock price
• Built a random forest classifier and a regressor to predict life time value of a new gambler from 1 month of interactions

Jul 2015 – Feb 2016 · 7 mos · Paris Area, France

• Built a Analytics Platform with batch mode/on-demand data ingestion from platform and Google analytics
• Built a reporting platform in Tableau and ETL pipeline in python
• Built a Monte Carlo simulation model for designing most profitable products

Dec 2013 – Jan 2015 · 1 yr 1 mo · Amsterdam Area, Netherlands

• Built models to predict market direction for various time horizons and analysed the evolution of relationship between covariates, historical time slice and prediction window

Apr 2010 – Dec 2012 · 2 yrs 8 mos · New Delhi Area, India

• • Built micro segments on students data (demographics, course data) to optimise ROI on campaigns

Dec 2005 – Mar 2008 · 2 yrs 3 mos · Bengaluru Area, India

• • Built diagnostic tool in C++ an Perl to automate bug discovery in the network by collecting data from switches, routers and storage devices