PyData London 2025 - Day1

PyData London 2025 - Day1

🚀 Tutorial Day at PyData London 2025!

Kicked off the conference with an amazing day of tutorials. Here’s what I dove into:

🧠 GPU Accelerated Python — by Jacob Tomlinson , Lawrence Mitchell , Jeremy Tanner and Katrina Riehl from NVIDIA

What I learned:

• How GPUs work (in a nutshell) ⚙️ • Writing Python code for GPUs with Numba 🐍⚡ • The ecosystem of GPU-accelerated Python tools like RAPIDS

🤖 Transformers Inside Out — by Sam Joseph

Key takeaways:

• A refresher on Transformer architecture 🧱 • Building a mini-GPT model from scratch using nanoGPT 🔧 • And yes… how to write good knock-knock jokes with AI 😄🪄

🕸️ Graph Theory for Multi-Agent Integration: Clinical Use Cases — by Ahmad Albarqawi

Highlights:

• Using graph-based architectures to orchestrate AI workflows 🧩 • Introducing IntelliNode, an open-source library 🧠💡 • Seeing MCP in action in real-world practice

All three sessions were incredibly insightful — I’ve already filled pages with notes and things to explore further! 📚📝

But learning must momentarily pause… because the next two days are packed with amazing talks! Can’t wait! 🙌🎤