Professional Experience

Network Guard

Aug 2018 – Present

Key Projects

Payments

I am the payments data specialist in Network Guard. I delivered data insights to help the payments team optimize their processes.

  • Fraud
    • Our payment fraud detection process had a rather high false positive rate – we were incorrectly classifying too many legitimate customers as fraudsters, which means revenue loss. I holistically reviewed the fraud detection process, identified key improvement areas and proposed recommendations which eventually resulted in a 4% decrease of false positive rate. 
      • I analyzed our fraud validation rules, identified those that had high false positive rates and liaised with developers to make those rules less aggressive. 
      • Some of our “high risk” customers are manually reviewed by our fraud operations team. However, the team doesn’t have a feedback loop in place, which means they could never learn from incorrect classification decisions in the past. I created a simple Tableau dashboard for the fraud operations team to review incorrect fraud classification decisions. Using a Tableau dashboard instead of requesting developers to build a web interface allows the team to start running the process sooner, hence realizing business value sooner.  
  • Operations
    • I created dashboards to visualize and monitor completion time of certain manual processes performed by the fraud operations team. The increased visibility has led to improved adherence to the completion deadlines. As a result, the manual processes are completed within the deadline over 95% of the time as opposed to 70% before the dashboard was implemented.
  • Product management
    • I took on additional responsibilities as product manager of an integration project with a payment processor during the absence of a product manager for the team. I specified the requirements, led user story mapping sessions and used data to quantify business value of backlog items in order to prioritize them. I also led bi-weekly demo sessions to key stakeholders of the project, which include the co-founders of the company.

User segmentation

I led a project to define and assign our users into different user segments according to their use cases. The user segmentation definition is now used company-wide by various business units to guide their customer acquisition and retention efforts.

  • We know what people use our services for but we have no idea how large each group is and how fast are they growing. We don’t have a model to segment our users by their use cases and this was made more challenging since we are a privacy-focused company and hence don’t collect many user data.
  • Based on the limited data we have, I came up with a simple user segmentation methodology and presented the results to key stakeholders to seek their feedback. After a few iterations, we were able to reach an agreed definition that will be used company-wide.
  • I also ran a workshop with various stakeholders to understand how insights related to user segments will be used across the company. This helped me prioritize dashboard and insights to share.

Emails Marketing Analytics

Metrics of our email marketing campaigns used to be scattered all over the place. Marketing managers had to access data on emails sent, click-throughs and conversions in different places and then manually compile the data in order to assess performance of our email marketing campaigns. I created a dashboard that enables marketing managers to track all email metrics in one place. They can also compare A/B tests results much more efficiently.

  • I worked with marketing managers to understand the objectives and success metrics of different email marketing campaigns.
  • I used Apache Airflow to combine data sources from Google BigQuery and AWS Redshift into one place. The combine data source is processed and updated daily, which is then visualized in Tableau.