Malek Senoussi

• Designed and optimized RAG pipelines: data ingestion, chunking strategies, embedding generation, vector search, and retrieval optimization for knowledge-grounded generation • Implemented LLM orchestration patterns with LangChain: multi-step reasoning, tool integration, API orchestration, and complex agentic workflows • Built production-grade evaluation framework (2K+ LOC, open-source): reusable components, observability, monitoring, and quality governance • Ensured scalability and reliability: systematic evaluation, regression testing, cost-efficiency analysis, and performance monitoring • Contributed to industrialization: established best practices, reusable architectural patterns, and documentation standards

• Designed and deployed end-to-end AI applications integrated into existing platforms and data systems (1M+ samples) • Built scalable, maintainable, and observable systems: data pipelines, quality monitoring, performance tracking • Integrated AI capabilities into business workflows: APIs, microservices, and enterprise platform integration • Collaborated with cross-functional teams (domain experts, data engineers, researchers) to identify and deliver high-value use cases • Ensured reliability and governance: validation protocols, documentation, reproducibility standards • Published in PLOS Computational Biology (2024) — production-quality research

• Contributed to industrialization of AI development: defined reusable components, frameworks, and best practices adopted by the team • Established evaluation and governance standards: mandatory experiment tracking (W&B), standardized benchmarks, reproducibility checklists • Enabled systematic adoption beyond prototypes: practices outlasted my tenure and became team standards • Managed 20,000+ GPU-hours: scalability planning, resource optimization, cost-efficiency