🚀 Bringing Explainability to AI Projects: From Black Boxes to Inspectable Systems

Workshop Co-organized by GDG Warszawa in collaboration with SoftServe

This community-driven event is designed for developers, tech enthusiasts, and innovators who want to dive into the latest trends and practical applications of GenAI and Google AI products.

LLMs are often treated as magic boxes that generate code, predictions, summaries, images, and business decisions. But as AI systems become more powerful, questions of trustworthiness, reliability, guardrails, and alignment become critical—not just nice-to-have vocabulary for strategy slides.

In this workshop, we will explore how explainability can help us understand, debug, and improve AI systems. We will briefly review common explainability approaches across different model types, then focus on mechanistic interpretability: looking inside open-weight models such as Gemma to understand what happens inside the neural network.

Through a practical use case, we will discuss how explainability can support accuracy, safety, alignment, and better engineering decisions. The goal is to help participants identify where explainability can be applied in their own AI projects immediately—not as academic decoration, but as a practical tool for building more trustworthy systems.

🧠 What you will learn

You will get a practical introduction to:

• why explainability matters in modern AI projects,

• how to debug and inspect model behavior,

• what mechanistic interpretability is,

• how open-weight models like Gemma can be explored,

• how explainability supports trust, safety, alignment, and better engineering decisions,

• where to apply these techniques in real-world AI systems.

🗓️ Agenda

17:00 – 17:15 — Doors open & check-in
17:15 – 17:30 — Welcome and introduction
17:30 – 20:00 — Workshop: Bringing Explainability to AI Projects
20:00 – 21:00 — Networking & pizza time 🍕

👨‍🏫 Speaker

Damyr Hadiiev is a Principal R&D Engineer with over 15 years of experience in software engineering, solution architecture, and AI-driven innovation. With an academic background in theoretical physics, he combines deep expertise in AI/ML, deep learning, GPU acceleration, HPC, physical simulation, and quantum computing.

He has a strong track record of building and mentoring high-performing research and engineering teams, as well as delivering production-grade AI systems. His work spans healthcare, life sciences, financial services, and advanced R&D domains, where he focuses on turning complex technologies into practical, scalable solutions.

✨ Why join?

Because building AI is not only about making models work — it is about understanding why they work, when they fail, and how to make them safer and more reliable.

Come for the AI insights, stay for the community, conversations, and pizza.

🍕 Networking & pizza included!

What to Bring

• Laptop (Required): You must bring your own machine to participate.

• Charger (Recommended): Ensure you have power for the duration of the event.

Important Notes

• Attendance: If you cannot attend, please cancel your spot. "No-shows" without cancellation may be blacklisted from future GDG Warsaw events.

• Community: Organized by GDG Warsaw in collaboration with SoftServe

Let’s Build with AI.