Code Along with ZAKA: Hands-On AI for Every Skill Level

A Hands-On AI Webinar Series by ZAKA

Artificial intelligence is evolving fast and ZAKA’s latest webinar series proves that learning AI doesn’t have to be passive. Across five immersive sessions, participants explored everything from medical imaging to no-code automation, fine-tuning large language models, and building smarter development workflows. Whether you’re a seasoned engineer or a curious student, this series offered something for everyone.

Episode 1: Detecting Kidney Stones with YOLOv10

The series kicked off with a deep dive into computer vision, spotlighting the YOLOv10 architecture and its application in medical diagnostics specifically kidney stone detection.

What participants learned:

How to set up a Google Colab environment for object detection
Dataset preparation: loading, preprocessing, and splitting
Fine-tuning a pre-trained YOLOv10 model on custom medical data
Real-time performance monitoring and deployment strategies

This session bridged the gap between AI and healthcare, attracting data scientists, ML engineers, and medical professionals alike.

Attendees left with practical skills and post-webinar resources to continue experimenting on their own.

Episode 2: Fine-Tuning Llama 3-8B for Domain-Specific Intelligence

Led by Ashraf Kaasamani, this session explored the art of customizing large language models. Participants got hands-on with the Llama 3-8B model, transforming it into a domain-specific assistant using the Yahma/Alpaca-cleaned dataset.

Key takeaways:

Setting up Colab for language model training
Strategic dataset preparation for optimal learning
Fine-tuning techniques for instruction-following behavior
Evaluating model predictions and refining outputs

This episode resonated with developers, students, and researchers eager to push the boundaries of contextual AI.

It wasn’t just about code it was about crafting intelligence that understands nuance.

Episode 3: No-Code Automation with Make.com

For those looking to simplify their workflows, Ashraf returned to host a session on Make.com a no-code platform that empowers users to automate tasks without writing a single line of code.

What was covered:

Navigating the Make.com interface
Building automations across multiple apps
Optimizing workflows for efficiency
Live Q&A for real-time troubleshooting

This session was a hit among entrepreneurs, professionals, and students who wanted to build smarter systems without diving into code.

Attendees received sample workflows and a quick-start guide to keep the momentum going.

Episode 4: Fine-Tuning DeepSeek-R1 with LoRA Techniques

Jean-Pierre Fakhry led a technical deep dive into fine-tuning DeepSeek-R1, a 1.5B parameter language model. Using LoRA (Low-Rank Adaptation), participants learned how to adapt large models efficiently.

Highlights included:

Step-by-step setup in Google Colab
Dataset loading, preprocessing, and splitting
Monitoring training metrics and evaluating predictions
Deployment strategies for real-world use cases

This session was ideal for ML engineers and educators looking to gain hands-on experience with cutting-edge fine-tuning techniques.

The interactive Q&A wrapped up the session with insightful exchanges.

Episode 5: AI-Powered Development with Cursor

Sary Mallak closed the series with a powerful session on integrating AI into software development workflows. This webinar showcased how Cursor AI and the OpenAI API can act as intelligent co-pilots for modern developers.

What participants built:

Real-time audio transcription using Whisper
GPT-powered RAG workflows for smarter responses
Functional interfaces using natural language prompts

Designed for developers and tech enthusiasts, this session emphasized practical implementation.

Attendees left with the tools and confidence to bring AI into their daily coding routines.

Final Thoughts: Learning by Doing

Each webinar in this series emphasized hands-on learning, collaborative exploration, and practical outcomes. From medical imaging to language models and automation, ZAKA’s sessions proved that AI education can be both rigorous and accessible.

Whether you attended one session or all five, the series offered a roadmap for anyone looking to deepen their understanding of AI and apply it meaningfully in their field.

Ready to join the next wave of AI innovation? Stay tuned for future webinars and keep building.