DSC MENA 2025
who it's for
data scientists, machine learning engineers, ai product managers, data engineers, security engineers, fintech leaders
topics
medical imaging with UNet and GANs; physics-informed neural networks for therapy; open source large language models; retrieval augmented generation systems; prompting and prompt engineering practices; agentic AI and automation pipelines; generative diffusion models; genai enhanced digital twins; payments fraud detection with AI; credit risk modeling with Kedro pipelines; setting up Iceberg tables in Snowflake; experimentation and controlled A/B testing
| title | author | length |
|---|---|---|
| How AI Is Transforming Medical Imaging | Svetlana Krunic | 29m |
Shows concrete results: UNet and GANs reduce noise and segment calcifications so that an original mammogram reviewed as cancer by 7 of 14 radiologists becomes… | ||
| Bias, Explainability, and Accountability: The Data Scientist's Burden | Youssef Kandil | 30m |
Examples of real failures (Tesla crash 2018, Apple card gender bias, COMPAS showing 23% vs 45% recidivism) motivate concrete fixes: prefer decision trees or… | ||
| Learn Like a Leader, Prompt Like a Pro: A Very Virtuous Cycle | Katherine Munro | 30m |
Claims that leadership habits map directly to prompting and provides a checklist plus concrete prompts: use Swisscom GPT to debug a voice bot (find where the… | ||
| Data Without Direction: The Cost of Ignoring the Business Model | Mohamed Fetiha | 22m |
Claims that most developed models never reach production, citing 70-75% of models shelved; shows a chatbot rollout that lowered satisfaction because root cause… | ||
| FinTech Startups vs Traditional Financial Institutions | Mohamed F., Basma R., Rana K. El S., Asmaa Magdy | 36m |
Panel documents concrete wins: a customer segmentation and recommendation engine for multiple lending products that drives targeted offers, a volatility… | ||
| From Risk to Resilience: How Security Can Shape the Future of Fintech | Nour A. Eldayem | 22m |
Security positions itself as an enabler for open banking, embedded finance and AI-driven services by reducing risk to an acceptable level rather than… | ||
| Beyond Transactions: Leading Fintech with Data, AI, and Monetization | Asmaa Magdy | 20m |
Practical roadmap for turning fintech data into repeatable revenue: a Klivvr case uses a quick POC delivered in weeks to prove value, selects use cases by… | ||
| AI, Metaverse & Blockchain: Shaping Digital Realities | Suren M. | 25m |
Platform-level product combining immersive worlds, on-chain identities and agent automation: deployed Metaverse presence at Emporio Armani Milan and an NFT… | ||
| Cloud Powered Fintech: The Future of Financial Services | Michael Moufeed | 29m |
Huawei Cloud launches an Egypt availability zone (May 2024) delivering ~26 ms latency locally and up to ~50 ms across North Africa, offers ~74 local services… | ||
| AI-Driven Strategies for Financial Innovation | Rana Kamal El Sharqawy | 18m |
Shows how EFG Holding builds a centralized data warehouse and data lake with ETL over six years, then layers business-driven AI to cut decision latency from… | ||
| Banking Transformation with AI: Extracting Value and Pioneering Change | Amira A. | 31m |
Claims concrete wins: banks report up to an 85% reduction in manual processes and roughly $3.2 billion prevented in fraud using real deployments. Practical… | ||
| AI-Powered Research: Accelerating Public Sector Transformation in the MENA Region | Medhat Kandil | 17m |
Generative workflows at Reso combine fine-tuning, retrieval-augmented generation and agentic AI to convert scattered sources into actionable research products… | ||
| The Future of Healthcare – AI’s Promise and the Challenges Ahead | Ashraf Bacheet, Mark Hanna | 28m |
Claims that AI already detects radiology abnormalities faster via radiology ML and that specialty models beat generic LLMs informally reported; teams use… | ||
| Open Source LLMs — Powering AI Accessibility and Innovation | D Bouabdallah, S Zaghloul, K Sunshev | 42m |
Open-source LLMs promise trust through inspectable weights but bring real risks: privacy, hidden engineering costs, and jailbreak/backdoor concerns. Panelists… | ||
| Data-Driven Transformation: Scaling Businesses with AI, Ethics, and Strategic Vision | D Bouabdallah | 14m |
Claims a retail AI system built with privacy-preserving machine learning (federated learning and differential privacy) plus a robust data governance layer… | ||
| AI for Good: Transforming Industries with Ethical & Innovative Solutions | Nourhan Fooda | 33m |
MENA stats: 37% of businesses implement AI in 2024, $320M went to regional AI startups in 2023, and 67% of executives flag ethics as an adoption barrier… | ||
| Battling the Machine: Cutting-Edge Attacks and Defenses in Generative AI | Mohammad Essam | 30m |
Examples show GPT‑3/GPT‑3.5 yielding two real Windows 10 activation keys exposed after web-sourced training data and a jailbreak that produced a poison recipe… | ||
| How RAG is Redefining AI Capabilities | Dina Helmy | 21m |
Shows how Retrieval-Augmented Generation is built end to end: document ingestion (offline) and retrieval (online) with parsing, chunking, embeddings and a… | ||
| Turning Data into a Strategic Asset | W Hossam, M Fouad, N Eldin Zoghly, A van der Haar, A El-Shafie | 30m |
Argues that companies turn data into revenue by monetizing telecom and sensor logs and avoiding GDPR fines through tighter governance, shows real scale like… | ||
| Agentic AI: What It Is and Why It Matters | Omar Sedeek | 29m |
Agentic AI augments LLMs with explicit reasoning, planning, memory and programmatic tool calls so models can book flights, run CI, write tests, and… | ||
| Vision, Big Data and AI | Hani Elgebaly | 28m |
AvidBeam converts real time surveillance into metadata and alerts and claims it scales to tens of thousands of cameras, citing city fleets like London with… | ||
| From Raw Data to Real Impact: Lessons from Real-World Transformations | Amr Saleh | 31m |
Telecom migration to a Snowflake cloud platform cuts a critical government report from three days to about 30 minutes and a one‑time large‑cluster test shows… | ||
| Transforming the MENA Region with GenAI Enhanced Digital Twins & the IOE | S. Adib | 21m |
Presents a concrete system: a hybrid physics–ML digital twin that updates online for real‑time structural health monitoring and aims toward an autonomous… | ||
| Traditional Governance Dead: How AI and GenAI Are Revolutionizing Governance | Ahmed El-Shafie | 29m |
Claims that AI and GenAI can auto-tag massive catalogs, infer lineage, and enforce privacy in near real time, backed by connectors to SQL Server, Oracle… | ||
| Data-Driven Culture in the Era of AI: What Still Matters? | Mirna Malaeb | 23m |
AI speeds insight generation but magnifies cultural gaps: decisions that act on AI outputs require a consolidated data platform, governance and human… | ||
| Hypothesis to Impact: Mastering the Art and Science of Controlled Experiments | Wael Aboulezz | 34m |
Controlled experiments prove that only about 20–30% of ideas succeed and learning beats guesswork: define a clear hypothesis, randomize by user into A/B, and… | ||
| AI & Data Science: Driving Innovation or Guarding Digital Borders? | Mohamed Fahmy, Mohamed Nabil, Mrs.Somaya El Sherbini, Ahmed Abdelhamid | 36m |
Panelists lay out concrete stakes: MENA AI market is about $11B now and projected to $150B by 2030, analytics AI drives ~75% of corporate use versus ~25%… | ||
| AI-Powered Strategy in the Age of Innovation | Laila Kakar | 26m |
Claims specific numbers: 1,500+ GenAI/ML tools, 10+ new LLMs weekly, 100+ open-source projects monthly, and 100+ AI startups monthly; a PwC survey reports 56%… | ||
| Harnessing the Power of LLMs in Advanced Analytics | Ahmed Abulkhair | 24m |
Covers concrete LLM history and numbers (GPT2 1.5B parameters, GPT3 ~175B trained on ~3,000 years of text) and practical builds: early text-to-SQL at ~50–60%… | ||
| The Official Opening | Amr Talaat, Peter Mollema, Walid Saad, Haitham Hamza, Aleksandar Linc Djordjevic, Amani Eissa | 37m |
Opens with a minute of silence for Dr. Reem Bahgat then lists concrete outcomes: over 100 speakers, 200+ hours of content and 40+ industry partners, plus a… | ||
| The AI Slingshot: How Small Teams Beat the Big Guys | Wael Salloum | 36m |
Small teams win by turning AI into fast product loops: claim that tasks once needing five experts and a year (sentiment model) now deploy in 30 minutes versus… | ||
| RAG How To Talk To Your Data | Sohaila Diab | 01h 33m |
Explains that Retrieval-Augmented Generation pairs LLMs with external sources to cut hallucinations and stay current and shows a working pipeline: Cohere LLM +… | ||
| Generative AI with Diffusion Models | Ahmed Ezzat | 01h 21m |
Walks through denoising diffusion probabilistic models with the concrete claim that diffusion now outperforms GANs on image synthesis, explains the forward… | ||
| Integrating AI into Your Apps: A Developer's Guide to Generative AI APIs | Hany Saad | 01h 38m |
A step-by-step developer demo of how to integrate Generative AI into apps using OpenAI Chat Completion API, showing code that builds an HTML chat page and a… | ||
| Multi dimensional audience segmentations in products | Aliaksandr Zhastkou | 50m |
Yango Delivery combines behavioral, geographic and ML signals to produce usable segments: classic RFM and an NPS+value map, an ML scoring feed that selects… | ||
| Inclusive & Ethical NLP Design Building AI for Everyone | Eve L. Parker, Hannah Claus | 01h 26m |
Workshop explains concrete methods: health‑first accessibility checks, WCAG/UAAG compliance, and iterative QA per task; an immersive avatar interface raises… | ||
| How to Construct and Assess Your First LLM Based RAG System | AbdElRhman ElMoghazy | 01h 34m |
Builds a working RAG chatbot that scrapes the Hugging Face NLP course into 95 documents, chunks with a recursive splitter, embeds with MiniLM L6 v2 embeddings… | ||
| Building Efficient and Frugal Image Data Processing Pipelines | Anna Chi | 01h 04m |
Pipeline recipes and measured results: switching PNG to WebP saves ~60% storage and moving 10 TB from S3 Standard to Glacier cuts a monthly bill from ~$300 to… | ||
| AI in Finance From Insight to Action | Manoj Tyagi | 01h 30m |
Builds two practical pipelines in Python using scikit-learn and pandas: a monthly‑expense predictor and an unsupervised fraud finder. The expense demo shows a… | ||
| AI Powered Uplift Modeling and Stream Forecasting in Music Industry | Maksim Karyagin | 01h 00m |
Shows concrete wins: uplift modeling cuts cost per add-to-library by ~30% using two-model, class-transformation and dedicated uplift random forest methods on a… | ||
| AI vs AI: Cyber Defence and Managed Resilience | Ahmed Elhaddad | 01h 19m |
Argues that cyber vendors move to cloud-first because hardware supply shocks produced a 400-day device backlog, cloud sandboxes and EDRs let vendors offload… | ||
| AI Advantage Reinventing SME Operations | Samitha Nagasinghe | 01h 16m |
Claims a 35% productivity uplift and shows two implementation paths: buy-in Azure bots for quick integration or self-host with Docker + Ollama + DeepSeek 7B on… | ||
| Automating Podcast Localization via OpenAI APIs | Rustam Musin, Grigory Panov | 01h 04m |
Demonstration shows an end-to-end proof of concept that pulls episodes via podcast4j / podcastindex.org, transcribes with Whisper-1, cleans and translates text… | ||
| AI Coding Agents and How to Code Them | Alex Shershebnev | 01h 28m |
Live coding builds AI coding agents that run a REACT-style thought→action loop, use Model Context Protocol servers (bash, filesystem, fetch, Brave search) and… | ||
| Real-Time Structural Health Monitoring Using Physics-Driven Digital Twins | Shady Adib | 01h 23m |
Claims a working hybrid digital twin that fuses physicsbased models with deep learning to do real‑time damage identification and short‑term prediction, uses… | ||
| AI Powered Fraud Detection for Payments & Transactions | Maimoon Saleem | 35m |
Claims that AI detects fraud in real time using anomaly detection (autoencoders, clustering), supervised predictive modeling (logistic regression, gradient… | ||
| From RAG to Riches Building Enterprise Search Systems with LLMs | Muddassar Sharif | 01h 28m |
Claims a production RAG pipeline is feasible: document parsing, semantic chunking with 10–30% overlap, OpenAI/Claude embeddings stored in a vector DB, and… | ||
| Experimentation in BigTech Validating Hypotheses and Driving Innovation | Parth Menon | 27m |
Meta-style experimentation shows how a single APK with Firebase Remote Config and Meta’s internal Quick Experiments framework runs gated A/B tests, sampling… | ||
| Evaluating RAG Model Performance Metrics, Bias, and Interpretability | Amir Siddiqui | 15m |
A concrete RAG pipeline is built end to end: chunking text with tunable chunk size, creating embeddings, storing vectors in Chroma and Pinecone, retrieving… | ||
| Methodology of Credit Risk Modeling Using Kedro Pipeline | Alexey Khoroshilov | 37m |
Argues that using the Kedro template plus Docker and MLflow logging turns ad‑hoc PD model work into a reproducible pipeline: a seven‑folder project (conf… | ||
| Setting Up Iceberg Tables in Snowflake | Ved Prakash | 52m |
Step‑by‑step recipe for running Apache Iceberg on Snowflake with code repo promised (lab failed live) showing how to: setup Snowflake external Iceberg tables… | ||
| The intersection of Quantum Computing and Generative AI GenAI | Sameh Zaghloul | 01h 35m |
Argues that hybrid quantum‑classical GenAI is already practical: a working QuantumGeni MVP uses a quantum SVM module run on a 128‑core simulator to classify… | ||
| Stay Ahead in AI with Product Engineering and AI Native Tools | Basalt Mounir | 01h 24m |
Framework that turns a monthly market‑research workflow into a modular AI pipeline: scrape headlines (News API or crawlers), clean text, produce embeddings via… | ||
| Task management for Data Professionals Ensuring clarity, timeliness & Impact | Natalia D. | 01h 25m |
Workshop shows concrete rules for task-ready data work: use SMART targets (e.g., increase model accuracy to >75% by May 1), write a CSD definition of done… | ||
| nothing matches | ||