Recommender Systems Specialist
CurrentByborg is an online entertainment company providing a global live-streaming platform in the adult industry (50M daily users). My main responsibilities are managing and developing the recommender system that powers the site:- Applied ML: frame the problem, design solution, conduct data and feature discovery, offline evaluation, A/B testing, continuous improvement and fine-tuning, stay updated with the state-of-the-art advancements.- New algorithm: design and implementation of a… Show more Byborg is an online entertainment company providing a global live-streaming platform in the adult industry (50M daily users). My main responsibilities are managing and developing the recommender system that powers the site:- Applied ML: frame the problem, design solution, conduct data and feature discovery, offline evaluation, A/B testing, continuous improvement and fine-tuning, stay updated with the state-of-the-art advancements.- New algorithm: design and implementation of a new model able to detect and recommend the most trending and engaging live streams, increasing user conversion.- Multi-list interface: led the design and implementation of a new interface with multiple variegated carousels, enhancing user acquisition.- ML framework ownership: prototype models, feature selection, offline evaluation and optimization (Python, Pandas, MLFlow, Optuna, Dagster, Evidently).- Analytics improvement: identification of new RecSys-related and business KPIs, architectural design of data pipelines, led the POC of an internal solution to enhance decision-making processes (Python, Dagster, DuckDB, Ibis, Superset)- Other: real-time and sequential recommendations, behavioural testing, and real-time image recognition to improve personalization. Regularly presented findings and strategies to stakeholders, driving key product decisions. Show less