Samira El-Masri

Design a comprehensive API gateway for LLM services with routing, rate limiting, authentication, observability, and provider normalization.

Build a comprehensive prompt regression testing suite with golden datasets, automated evaluation, and CI/CD pipeline integration.

Comprehensively analyze embedding space quality through structural, semantic, and retrieval analysis with visualizations and improvement recommendations.

Implement LLM self-consistency checking with multiple generation, contradiction detection, and consensus selection methods.

Build a scalable document deduplication pipeline using hash, MinHash/LSH, and embedding-based methods with merge strategies.

Implement comprehensive error budget management for AI systems with SLO frameworks, burn rate alerting, and policy enforcement mechanisms.

Implement retrieval diversity optimization using MMR, clustering, and aspect-based methods with quantified quality-diversity trade-offs.

Build an LLM confidence calibration system combining multiple signals with trained calibration curves and reliability metrics.

Build a comprehensive load testing framework for LLM APIs with multiple scenarios, detailed metrics collection, and automated reporting.

Design a complete multimodal RAG pipeline supporting text, images, and documents with cross-modal search and vision-language model integration.

Systematically analyze RAG failure modes from production data, categorize issues, identify root causes, and propose quantified mitigation strategies.

Design comprehensive negative sampling strategies for retrieval training with easy, medium, and hard negative mining and curriculum learning.