Badr Lakhal

– Consulted for L'Oréal within the BPMa team, maintaining and optimizing a data platform while driving performance improve- ments, data quality initiatives, and new feature delivery. – Engineered a data-cleaning logic that reduced production tables by 47%, cutting storage costs and maintenance overhead. – Resolved dashboard performance bottlenecks by implementing precision optimization on large numeric columns, reducing report load times.

– Migrated EC2 cron jobs to serverless AWS Lambda/EventBridge, reducing idle compute costs by 93% and improving uptime. – Built an agentic AI system for a UK/Germany client, reducing manual sales research time by 25%. – Built an AI system for candidate screening, exposing LLM-driven workflows via FastAPI and Docker, with vector-based retrieval and automated testing.

– Built a scalable ETL pipeline to ingest, clean, and transform unstructured documents, doubling the volume of usable training data for model development. – Developed and deployed a Naive Bayes classifier for automated document screening, reducing manual review effort and processing time. – Performed statistical analysis on processed datasets to identify early data quality issues and optimize downstream processing workflows.