root@avigdorsky:~$ neofetch
██████╗ ██╗ ██╗ avigdor@avareange
██╔══██╗██║ ██║ ---------------------------------------------
███████║██║ ██║ Role ......: Co-founder @ AVAREANGE · full-stack dev & researcher
██╔══██║╚██╗ ██╔╝ Coding ....: since 2008 · full-stack all the way
██║ ██║ ╚████╔╝ Focus .....: AI agents · ML / CV · cybersecurity
╚═╝ ╚═╝ ╚═══╝ Also ......: CTO @ CyberSaver (cybersaver.io)
Education .: Ph.D. (Eng.) · MIPT (Phystech) · MEPhI (cybersecurity) · Tsinghua
Location ..: Tel-Aviv (Arctic-born)
Shell .....: zsh + LangGraph + Telegram
Uptime ....: 5 countries · 3 startups · ∞ coffeeReduce human cyber risk with AI agents. People are the #1 attack surface. AVAREANGE is an AI-first platform that gives companies a virtual CISO, continuous human-risk scoring, phishing & social-engineering simulations, GRC/SOAR/SAT — in one OS.
a1b2c3d 2008 first lines of code — self-taught from books, before owning a computer
b2c3d4e MIPT Moscow Institute of Physics and Technology (Phystech)
c3d4e5f MEPhI National Research Nuclear University — cybersecurity
d4e5f6g Tsinghua postgraduate research, Beijing
e5f6g7h Ph.D. Candidate of Engineering Sciences
f6g7h8i now Co-founder @ AVAREANGE · CTO @ CyberSaver · full-stack dev & researcher
🛡️ AVAREANGE — Human Security OS prod
Turn employees into the first line of cyber defense in 2 weeks. Listed in the Russian Unified Software Registry. Customers report 5× fewer human-caused incidents and −70% attack risk.
| Module | What it does |
|---|---|
| 📊 Human-risk scoring | Behaviour-based risk model per employee / department |
| 🎣 Phishing simulations | Automated pentest of people: email, attachments, messengers |
| 🎓 LMS + monitoring | SCORM courses, real-time dashboards and analytics |
| 🤖 AI phishing analyst | Chat-bot, mail plugin and browser extension that scores suspicious content |
| 🏢 Org & access | Org hierarchy, LDAP / AD / SSO / 2FA integrations |
| 🧩 MSSP panel | Multi-tenant console for managed-security partners dev |
🎮 CyberSaver / Киберзащитник — pet project, CTO active
Gamified mobile app that makes anyone a conscious, protected user: AI assistant that scores suspicious messages & links (1–10), 3–5-minute courses, mini-games, XP · badges · leaderboards. iOS · Android · RuStore. Swift · Kotlin · LLM agents.
ai_agents:
runtime: [LangGraph, LangChain, MCP servers & tools, multi-agent orchestration]
patterns: [agentic RAG, graph-RAG (Apache AGE + pgvector), tool-use, long-term memory, evals, guardrails]
models: [self-hosted LLMs via vLLM, Qwen, DeepSeek, GigaChat, Claude, GPT]
ml:
frameworks: [PyTorch, HuggingFace Transformers, scikit-learn, sentence-transformers]
cv: [OpenCV, YOLO / detection & tracking, OCR pipelines, image classification]
nlp: [phishing / social-engineering classification, human-risk scoring, anomaly detection, RU-language NLP]
ops: [fine-tuning & LoRA, embeddings, evaluation pipelines, on-prem inference]
full_stack:
languages: [Python, TypeScript, Go, Swift, Kotlin, Bash]
backend: [FastAPI, Node.js, event-driven microservices, REST / gRPC / WebSockets]
frontend: [React, Next.js, TypeScript]
mobile: [Swift / iOS, Kotlin / Android]
data: [PostgreSQL, Apache AGE (graph), pgvector, Redis, Kafka]
messaging: [Telegram Bot API, MTProto, webhooks, browser extensions, mail plugins]
security:
networks: [TCP/IP, TLS, DNS, packet analysis (Wireshark), network hardening, VPN topologies]
offense: [phishing & social-engineering simulations, OSINT, red-team automation]
defense: [threat intel, malware / exploit analysis, detection engineering, human-risk analytics]
appsec: [OWASP ASVS / MASTG, web & mobile security, reverse engineering basics]
devsecops: [Docker, Kubernetes, GitHub Actions, self-hosted Gitea inside VPN, Nginx, Linux]
+ shipping AVAREANGE — AI phishing analyst & human-risk scoring
+ building MSSP client panel · CyberSaver mobile app
+ exploring agentic RAG over graph + vectors (Apache AGE + pgvector)
+ reading Human Risk Management · cognitive science · threat intel
! skeptical of hype — speed, leverage, execution
auto-updated daily via GitHub Actions