# ~/.config/identity.pyname="Nikan Eidi"location="Toronto, CA · UTC-5"status="ONLINE — open to work"languages= ["Python", "Swift", "TypeScript",
"Java", "Bash", "SQL", "C/C++"]
ml_ai= ["PyTorch", "scikit-learn", "MLX",
"Gemma 2/3", "LangChain", "ONNX"]
systems= ["SwiftUI", "MVVM", "Vision FW",
"AVFoundation", "Metal / MPS"] | # ~/.config/stack.pycloud= ["AWS · S3 · DynamoDB",
"Cognito · EC2 · ECR",
"Docker", "GitHub Actions", "n8n"]
frontend= ["React 19", "Vite", "Flask",
"Streamlit", "Supabase",
"Framer Motion", "Tailwind"]
mantra= ["architecture > complexity",
"ship fast, scale elegantly",
"build for humans, not specs"] |
%%{init: {'theme':'base','themeVariables':{'primaryColor':'#0d0226','primaryTextColor':'#F0F5FC','primaryBorderColor':'#8426FF','lineColor':'#00D8F2','fontFamily':'monospace'}}}%%
flowchart TB
CORE(("⬡ NIKAN\nEIDI"))
CORE --> NATIVE["📱 NATIVE & SYSTEMS"]
CORE --> ML["🧠 ML / AI"]
CORE --> CLOUD["☁️ CLOUD & FULLSTACK"]
CORE --> TOOLS["⚙️ DEV TOOLS"]
NATIVE --> HC["HC — Forge Dragon Terminal\nSwiftUI · Voice NLP · Vision"]
NATIVE --> VN["VisionNotebook\nSwiftUI · AppKit · Spatial UI"]
ML --> SN["StarryNote\nMLX · Gemma 3 · 382 tests"]
ML --> DC["DeepCIFAR\nResNet18 · 96% acc"]
ML --> AU["Aurora Kraken\nasyncio · OSINT · Gemma 2"]
CLOUD --> FR["Fragments\nNode · AWS · CI/CD"]
CLOUD --> CV["CineVision\nReact · Flask · TF-IDF"]
TOOLS --> AX["AX\nBash · GPT-4o · 16fps TUI"]
TOOLS --> GS["Git-Sync\npure Bash · ANSI UI"]
TOOLS --> GV["GitVisionCLI\nMulti-model AI IDE"]
classDef core fill:#8426FF,stroke:#00F273,stroke-width:3px,color:#fff
classDef cat fill:#0d0226,stroke:#00D8F2,stroke-width:2px,color:#00D8F2
classDef proj fill:#020204,stroke:#8426FF,stroke-width:1px,color:#F0F5FC
class CORE core
class NATIVE,ML,CLOUD,TOOLS cat
class HC,VN,SN,DC,AU,FR,CV,AX,GS,GV proj
Cyberpunk work-hour tracker on the Forge design system. Continuous voice command engine (Apple Speech + NaturalLanguage), front-camera hand-gesture control with One-Euro filter, and an ANSI dragon terminal rendered per-character. | |
|
Local-first AI study engine. Feeds any file — PDF, image, code, notes — through Gemma 3 on Metal GPU and synthesizes a 10-section study guide with original proofs, Mermaid diagrams, and exam questions. Triple-layer post-processor. 100% test pass rate. |
AI-driven intelligence pipeline. 12-layer identity resolution across 15+ concurrent async leads, speculative TLD probing, and a local AI Surgeon (Gemma 2-2B) that reads non-standard DOMs when regex fails. 150+ cached leads/sec. | |
|
Single-file Bash utility that fingerprints your stack and dispatches to GPT-4o for a production CI workflow — while a 16fps Cybernetic Eye animates in terminal via pure |
Full-stack tracker with a TF-IDF cosine-similarity engine over 10,000+ titles. Liquid Glass UI, Coverflow search, episode-level progress, and real-time watchlist-driven recommendations. | |
|
End-to-end CIFAR-10 classifier. Transfer learning with pretrained ResNet18 fine-tuned on Apple Silicon, AdamW + CosineAnnealingLR, full inference CLI. 96.02% test accuracy in 6 epochs. |
╔════════════════════════════════════════════════════╗
║ "Technology is not about solving problems — ║
║ it's about reimagining what's possible." ║
║ — Nikan Eidi ║
╚════════════════════════════════════════════════════╝