class DrDiffie:
"""Like Diffie-Hellman, but for knowledge exchange"""
background = ["Pharmacy", "Biochemistry"]
passion = ["Low-level ML Optimization", "Computer Vision", "Cybersecurity"]
current = "Automated Trading Systems & Neural Network Surgery"| 🚀 Active Projects | 🔬 Deep Dives |
|---|---|
| Automated Trading Systems ML-driven market analysis without traditional indicators |
llama.cpp Internals Low-level weight optimization & quantization |
| Computer Vision Pipeline Medical imaging & molecular structure recognition |
Cybersecurity Research CVE hunting & bug bounty programs |
| Biomedical NLP Extracting insights from pharmaceutical literature |
From-Scratch Implementations Building to understand, not just to use |
+ Recreated Monte Carlo Counterfactual Regret Minimization from first principles
+ No libraries, no shortcuts - pure algorithmic implementation
+ Because understanding > importing ⚗️ Chemistry Experiments
/ \
/ \
📚 👨🍳
Books Cooking
& |
Learning Creating
When not optimizing neural networks, you'll find me experimenting in the kitchen, diving into books, or conducting chemistry experiments (safely, of course!).
"Like the Diffie-Hellman key exchange, the best solutions come from combining seemingly unrelated domains into something cryptographically beautiful."
Open to collaborate on:
- 🧬 Medical AI & Biomedical Applications
- 🔐 Security Research & Vulnerability Discovery
- ⚡ Low-level ML Optimization
- 🎯 Algorithmic Trading Systems



