MenaML 2025 / Sponsored by Google DeepMind and QCRI

I had an amazing time at MenaML’25, the MENA Machine Learning Winter School sponsored by Google DeepMind, so I wanted to share some experiences and photos (a lot!) so you might also think about applying to MenaML. Before I start, let me share information about MenaML’26!

Here’s some information about MenaML 2026 if you need it: mena.ml; linkedin.com/company/mena-ml

Application Deadline: 18 September 2025 Programme Dates: 24 - 30 January 2026 Location: King Abdullah University of Science & Technology, Thuwal, Saudi Arabia

Details and the application form here: apply.mena.ml

The programme offers a unique chance for graduate students, researchers, and engineers to learn directly from global leaders. There will be tutorials, workshops, and inspiring panel talks covering key topics like:

LLM Training & Inference at Scale
AI for Sciences, Healthcare, Agriculture & Chip Design
AI in Social Sciences & Urban Planning
Ethical AI & Responsible Innovation

I will share in detail, but if you want to see the highlights from last year’s event, check here: 2025.mena.ml

Let’s start →

Day 0: The Journey Begins

I flew from Ankara to Doha with excitement. I was selected as a fully funded participant for MenaML. This conference was organized by Google DeepMind and QCRI. I would meet talented people from 37 countries.

Day 1: Opening & Keynote

The conference started with welcome talks. Maria Abi Raad and Sami Alabed from the MenaML team opened the event. Dr. Ahmed Elmagarmid from QCRI shared his vision for AI in our region.

Opening: YouTube - MenaML Opening

Then came the highlight. Shakir Mohamed gave the keynote speech. He’s the Director of Science at Google DeepMind. TIME magazine named him one of the 100 most influential people in AI. I had the chance to speak with him over lunch. Hearing about his journey and leadership in Deep Learning Indaba, an initiative spanning Africa, was truly inspiring.

Shakir Mohamed’s Talk: YouTube - Shakir Mohamed Keynote

Five Days of Learning

Sunday, February 9: We started with deep learning basics. Sahra Ghalebikesabi taught us the fundamentals. Ehsaneddin Asgari explained transformer architecture. In the afternoon, we had hands-on practice with LLMs. Ibrahim Alabdulmohsin and Mehdi Bennani (he was my poster’s judge :D) guided us through the exercises.

Introduction: YouTube - Deep Learning Basics
Transformers: YouTube - Transformer Architecture

Monday, February 10: Mohamed Eltabakh spoke about Arabic LLMs. Mohammad Raza taught us about LLM reasoning. Pranali Yawalkar led a practical session on training LLMs efficiently. We also had a career panel with regional AI leaders. The panelists included professors from HBKU and Carnegie Mellon, plus CEOs from QeenAI and PhazeRo.

Arabic LLMs: YouTube - Arabic Language Models
Reasoning Models: YouTube - LLM Reasoning

Tuesday, February 11: We learned about generative AI and diffusion models. Safa Messaoud and Andrew El-Kadi were our teachers. Andrew showed us how to use diffusion models for weather forecasting. Amin Sadeghi taught multi-modality. Badih Ghazi explained privacy issues in generative AI. I presented my poster on LLM agents. DeepMind engineers gave me valuable feedback.

Diffusion Models: YouTube - Diffusion Models
GenAI and Diffusion (GenCast is really interesting work): YouTube - GenCast
Multimodality: YouTube - Multimodal AI

Wednesday, February 12: This day focused on reinforcement learning. Sanjay Chawla introduced the basics. Mina Khan explained deep RL. Yasin Abbasi Yadkori taught us RLHF through hands-on practice. Chris Brown presented AI for Sustainability. We had another poster session in the afternoon.

Reinforcement Learning: YouTube - RL Basics
RLHF for LLMs: YouTube - RLHF

Thursday, February 13: The last day showed AI’s real-world applications. Experts presented AI in education, health, algorithms, biology, and genomics. Each talk showed how AI can change these fields.