Watch Recording
Enter your details to unlock
Free
Everything you need to build production-ready RAG pipelines from scratch.
Design a RAG Pipeline That Actually Works
Understand the retrieval-generation loop and why naive approaches fail
Learn chunking strategies, embedding models, and reranking techniques
Choose the Right Vector Database for Your Scale
Compare Pinecone, Weaviate, pgvector, and Chroma across latency, cost, and ops burden
See when you don't need a vector DB at all
Handle the Hard Parts: Evaluation and Hallucination
Build automated eval pipelines that catch hallucinations before users do
Implement citation grounding and confidence scoring
Ship RAG to Production with Confidence
Caching, batching, and cost optimization patterns that cut API bills by 60%
Monitoring and observability setup for retrieval quality over time
Every company sitting on internal docs, support tickets, or knowledge bases is a RAG use case waiting to happen. Yet most tutorials stop at "hello world" demos that break the moment you feed in real data.
This session covers the production patterns that separate toy demos from systems that handle 10K+ documents with sub-second latency. Whether you're building internal tools or customer-facing products, RAG is the bridge between your data and useful AI.
Ziyad has led design teams at some of the most impactful tech companies in Saudi Arabia and the broader MENA region. Starting his career at STC, he quickly rose to lead the UX redesign of their consumer-facing apps, serving over 20 million users.
At Noon, he established the company's first design system, reducing UI inconsistencies by 60% and cutting development time for new features by 40%. His work at Careem focused on the driver experience, where his research-driven approach increased driver satisfaction scores by 25%.
Currently consulting for NEOM's digital experience team, Ziyad brings a unique blend of regional cultural understanding and world-class design practices to every project he touches.
Worked at

Career highlights