RAG in Action: Building AI That Knows Your Data
Details
# RAG in Action: Building AI That Knows Your Data — Reem Zekry ·
---
الـ LLM ممكن يجاوبك بثقة… حتى لو مش عارف المعلومة الصح 👀
طب إيه اللي يحصل لما نخليه يدور في documents بتاعتنا، يختار المعلومات المناسبة، وبعدها يجاوب بناءً عليها؟ هنا بييجي دور الـ RAG 🔎
يوم السبت 31 أكتوبر، م. ريم ذكري معانا في Live session بعنوان RAG in Action: Building AI That Knows Your Data 🔥
هنبني الصورة خطوة بخطوة:
🔹 From documents to knowledge: إزاي نحمّل الـ documents ونقسّمها لـ chunks مفيدة
🔹 Embeddings & vector search: إزاي نحوّل النصوص لتمثيلات رقمية ونسترجع الأجزاء الأقرب للسؤال
🔹 Grounded answers: إزاي نوصل الـ retrieved context بالـ LLM عشان يجاوب بناءً على المعلومات المتاحة
🔹 Testing the pipeline: إزاي نقيّم جودة الـ retrieval، ونتعامل مع الإجابات الغلط والـ hallucinations
مش مجرد شرح نظري؛ هنفهم الـ RAG pipeline كـ system متكامل، من أول document لحد الإجابة النهائية.
م. ريم AI Engineer متخصصة في Data Science و GenAI، وAI researcher، وبتشتغل AI Architecture Consultant مع startups كتير. وزي ما بتقول عن نفسها: She cooks AI solutions.
لو بتسمع عن RAG كتير وعايز تفهم إزاي تبنيه فعليًا، السيشن دي ليك.
⏳ السبت 31 أكتوبر ٧ مساء بتوقيت القاهرة
📩 الرابط البث المباشر: https://www.facebook.com/GDGOctober/
📩 سجل من هنا: https://gdg.community.dev/e/mv5fpa/
اعمل mention لصاحبك اللي عايز ينقل الـ AI projects بتاعته من demo لحاجة بتشتغل على data حقيقية 👇
#GDG6October #RAG #GenerativeAI #LLM #GoogleDeveloperGroups #Egypt
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
An LLM can sound confident and still get the answer wrong. What if it could search your own documents first? 👀
On Saturday, 31 October, @Reem Zekry goes live with “RAG in Action: Building AI That Knows Your Data.”
We’ll walk through the pipeline behind a document-aware AI assistant:
🔹 Chunking documents into useful pieces
🔹 Using embeddings and vector search to retrieve relevant context
🔹 Giving that context to an LLM to generate grounded answers
🔹 Testing retrieval quality and understanding where hallucinations can still happen
The focus is practical: understanding how the pieces fit together, from the original document to the final answer.
Reem is an AI Engineer and researcher who consults on AI architecture for startups. Her own one-line bio? “She cooks AI solutions.”
If you want to move beyond calling an LLM API and understand how to build AI on your own data, this session is for you.
⏳ Sat, 31 Oct · 7:00 PM Cairo time
🔗 Live Link: https://www.facebook.com/GDGOctober/
📩. Registration Link: https://gdg.community.dev/e/mv5fpa/
Tag the friend who’s building their next AI project 👇
#GDG6October #RAG #GenerativeAI #LLM #AIEngineering #GoogleDeveloperGroups
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
A language model can generate a fluent answer without having the information your application actually needs.
RAG addresses this by retrieving relevant information from an external knowledge source and providing it to the model as context. But a reliable RAG system involves more than connecting a vector database to an LLM: document processing, chunking, retrieval quality, prompt design, and evaluation all matter.
On Saturday, 31 October, Eng. Reem Zekry will lead “RAG in Action: Building AI That Knows Your Data.”
The session will cover:
• Document preparation: loading, cleaning, and chunking source material
• Embeddings and vector search: representing text and retrieving relevant passages
• Grounded generation: supplying retrieved context so answers are based on the available evidence
• Testing and improvement: exploring retrieval quality, chunk size, top-k, and common failure cases
• Understanding limitations: why RAG can reduce some hallucinations but cannot guarantee every answer is correct
The goal is to understand the end-to-end RAG pipeline and how its design choices affect the quality of an AI application.
Eng. Reem is an AI Engineer focused on Data Science and GenAI, an AI researcher, and an AI architecture consultant for startups. She is also active in tech leadership and community empowerment initiatives.
If you're a developer, student, or AI practitioner looking to build applications that work with your own documents and knowledge sources, join us.
📅 Sat, 31 Oct · 7:00 PM Cairo time
🔗 Live Link: https://www.facebook.com/GDGOctober/
📩. Registration Link: https://gdg.community.dev/e/mv5fpa/
Know someone building with LLMs? Tag them.
#GDG6October #RAG #GenerativeAI #LLM #AIEngineering #GoogleDeveloperGroups
Agenda
---
Speaker
Reem Hatem
AI Engineer متخصصة في Data Science و GenAI
AI Researcher
AI Architecture Consultant لعدد من الـ Startups
Ambassador للـ Tech Leadership وتمكين الـ Communities في مبادرات وفعاليات تقنية متعددة
Host
Ahmed Fekry - Vodafone (iOS Developer)
Ahmed Fekry is a passionate iOS Developer at Vodafone and a certified Swift developer by Apple.
He currently leads the GDG 6 October chapter Lead and serves as an external instructor at the Information Technology Institute (ITI), where he also graduated from the prestigious 9-month Professional Diploma.With a strong foundation in app development, Ahmed participated in the Hajj Hac…
Complete your event RSVP here: https://gdg.community.dev/events/details/google-gdg-6-october-presents-rag-in-action-building-ai-that-knows-your-data/.
