How I built a 39× compression pipeline with AES-256-GCM in Python (and why the dictionary is everything)
I store LLM training data. Every tool I found either compresses it or encrypts it — nothing did both. So I built QUANTUM-PULSE.
The pipeline
payload → MsgPack → Zstd-L22 + corpus dict → AES-256-GCM →
quantum-pulse.hashnode.dev3 min read