Sardor Razikov
GitHub

Sardor Razikov

@SRKRZ23

View GitHub Profile

About

Independent AI/ML engineer · Cost-efficient & on-prem LLM infrastructure · AI governance.

Location Tashkent, Uzbekistan
Company Independent AI Safety Research
Followers 10
Following 4
Public Repositories 42
Expertise

Skills & Technologies

JavaScript Python
GitHub Work

Projects & Repositories

Recent public projects and repositories from this profile.

japan-mini-app

JavaScript

No project description available.

⭐ 0 Forks: 0
JavaScript View Project

lemonade-sec

Python

Private-by-design AI security code reviewer, powered by AMD Lemonade — SAST + local-LLM triage, your source never leaves the machine. MIT.

⭐ 2 Forks: 0
Python View Project

argus-aip

Python

Agentic supply-chain risk scanning, modeled as a live Foundry Ontology. Built for Palantir Startup Fellowship Cohort 003.

⭐ 0 Forks: 0
Python View Project

population-flow-reaction-sim

Python

Honest backtest: can LLM persona-reaction simulation (MatrAIx) reproduce a real, peer-reviewed flood-disclosure natural experiment? Direction: yes (5/5 models). Magnitude: overestimated ~2.5-3x.

⭐ 0 Forks: 0
Python View Project

agent-trust-safety

Python

Agent Trust & Safety tools: AI-agent config leak scanner (Apple CLAUDE.md-class) + OAuth blast-radius auditor (JPMorgan Opet-letter-class). MIT.

⭐ 1 Forks: 0
Python View Project

agentrank

Python

AgentRank — how easily is each frontier LLM hijacked by a hidden prompt injection in the web content it reads? Open, canary-scored, reproducible benchmark for AI-agent security.

⭐ 1 Forks: 0
Python View Project
← Back to Talent Open GitHub Profile ↗