Imad-Eddine M.

My tasks as Data Science Manager involve a mix of designing and implementing AI/Data solutions as well as establishing good ML practices.

Some projects I have worked on:

Design and development of a data drift reduction system leveraging context-enriched LLMs and clustering techniques, leading to a 30% improvement in recall and a 5% increase in conversion rates for black-box ML models.

Design of a creative asset evaluation system using VLLMs and statistical analyses for confidence and ratings.

Implementation of an end-to-end pipeline for CI/CT/CD of models using VertexAI & Kubeflow.

Design of an LLM-based topic modeling for share of voice.

Responding to ad-hoc requests related to customer segmentation modeling using analytics data.

During my year as a junior, I had the opportunity to:

Participate in NLP research projects and implement some of them.

Implement customer segmentation models and a Value Based Bidding system using GA4 data.

Conduct A/B tests and causal inference analyses evaluating campaign effectiveness on various KPIs.

Develop data processing pipelines for different ingestion systems.

Lead an exceptional team during the Jellyfish x Google Hackathon.

• Analysis of customer request and feedback flow

• Creation of dashboards and presentation of performance analyses

• Development of chatbots and analysis of requests processed by them.

Environment: Zendesk, Python