Instagram Reels Core Modeling.

Instagram Reels Core Modeling.

• Improve model architectures of instagram reels and explore by using state of the art architectures.

• Leverage LLM to improve recommendation quality.

• Improve model freshness/responsiveness.

• Improve model reliablity/stability.

Shopping Ads Personalization

• Improve shopping ads models (auction pCTR, factorized pCTR, Click Quality, ...)

• impact: increase revenue in order of 100 Million $

• Over 12 launches in a year

• proposed a lot of new ideas and directions

• successful collaboration with more than 5 teams

Fraud Transaction on Online Grocery and Online Return

• Develop and deploy and maintain machine learning model to capture online fraud transactions

• Impact: save around 10 Millions of $

• Develop and deploy and maintain pipeline of data for the related models.

• Actively working with business and stakeholders to give them insight about recent trends of fraud activities.

Research design implementation models that can work on non-stationary and evolving data.