Enterprise Storage Engineer × AI/GenAI Researcher · Bridging the infrastructure that makes AI possible

I sit at the intersection of two worlds most engineers only know one of — enterprise-grade storage systems and cutting-edge AI infrastructure. I've worked hands-on with IBM Storage, EMC PowerMax, and SAN fabrics at enterprise scale. Now I'm obsessed with how AI systems store, retrieve, and process information — and I'm building the open knowledge base that connects both worlds.

Storage is the foundation AI is built on. I know both sides.

identity = {
    "enterprise_storage": {
        "systems":    ["IBM Storage (FlashSystem, DS8000, Spectrum)", "EMC PowerMax"],
        "networking": ["Brocade SAN Switches", "Cisco MDS & Nexus", "FC/FCoE/iSCSI"],
        "concepts":   ["SAN Architecture", "NAS Design", "Storage Fabric Zoning"],
        "cloud":      ["Cloud Storage Integration", "Hybrid Storage Architecture"],
    },
    "ai_and_genai": {
        "current":    ["LLM Inference Systems", "KV-Cache Optimization", "vLLM", 
                       "Generative AI Architecture", "Vector Databases"],
        "storage_ai": ["GPU Memory Management", "Flash Attention", "Model Storage",
                       "Checkpoint Systems", "Distributed Training I/O"],
        "future":     ["Agentic AI Systems", "Multimodal Storage", 
                       "Neuromorphic Computing Storage", "AI-Native Storage"],
    },
    "building": {
        "storagecraft": "Open knowledge base — storage + AI infra",
        "simulators":   "Interactive visual learning tools",
        "benchmarks":   "Reproducible, Python-scripted real data",
        "mission":      "Make enterprise + AI knowledge free forever",
    },
}

The open knowledge base where enterprise storage meets AI infrastructure. Real knowledge. No paywalls. No NDAs. Open forever.

🌐 manmeetnain.github.io/storagecraft ⭐ github.com/manmeetnain/storagecraft

▶ RAID-5 Visualizer See parity distribution, disk failures, and rebuild in real-time

LSM Tree Simulator — Week 4 Attention KV-Cache Visualizer — Month 2 SAN Fabric Zoning Simulator — Month 2

• Write Amplification — The hidden cost multiplier in SSDs, LSM trees, and distributed storage
• Write-Ahead Log — How crash safety works in every major database
• KV-Cache in LLM Inference — The most critical storage optimization in modern AI

Generative AI ──────────────────────────────────────────
  How LLMs store and retrieve knowledge (KV-cache, RAG)
  Agentic AI memory systems and persistent storage
  Vector database internals (HNSW, IVF, PQ)
  Multimodal AI storage patterns

AI-Native Storage (the future) ─────────────────────────
  Storage systems designed for AI workloads from ground up
  Checkpoint-optimized filesystems for training runs
  Near-storage compute for ML inference
  Neuromorphic and in-memory computing storage
  
Enterprise × AI Intersection ───────────────────────────
  Running LLM inference on enterprise storage arrays
  IBM and EMC integration with AI pipelines
  SAN fabrics for GPU cluster interconnect

Whether you need enterprise storage architecture advice or help designing AI infrastructure — I bring both worlds together.

StorageCraft is free, open source, and always will be. Enterprise storage knowledge and AI infrastructure insights shouldn't be locked behind paywalls. If this work helped you — consider sponsoring to keep it going.

→ Become a Sponsor

Tiers from ₹200/month. Every contribution funds more deep dives.

Enterprise Storage × Generative AI × Open Source · Building from India 🇮🇳 · One commit every day