Ayyan = {
"name" : "Muhammad Ayyan Khan",
"location" : "Peshawar, Pakistan",
"role" : "Data Analyst & ML Engineer",
"education" : "B.S. Data Science (Current)",
"work" : "Freelance Data Analyst — Remote",
"building" : ["Git-Mind (multi-agent code analysis)", "Dep-Age (Rust CLI)"],
"learning" : ["Deep Learning", "NLP", "Rust", "Time Series Forecasting"],
"contact" : "ayyan9466@gmail.com"
}I'm a Data Analyst and ML Engineer who turns messy data into actionable insights. My work spans the full pipeline — from exploratory analysis and visualization to building production-ready ML systems with experiment tracking, data versioning, and deployment. Lately I've been diving into Rust to build fast developer tooling alongside my data/ML work. I believe the best solutions come from understanding both the math and the engineering.
I'm most excited when a model goes from Jupyter notebook to production pipeline — watching data transform into decisions is what keeps me coding. 🚀
Languages — Python for data/ML, Rust for performant tooling, SQL for databases
Data & Visualization — wrangling, analysis, and interactive dashboards
ML & AI — from classic sklearn to deep learning and LLM agent frameworks
Tools & Platforms — MLOps, APIs, containerization, and version control
- 🧠 Git-Mind — multi-agent code analysis platform (FastAPI + LangGraph + Rust)
- 🛡️ crate-security-auditor — Rust crate vulnerability scanner
- 📦 Dep-Age — CLI dependency age checker (Rust, npm, PyPI)
Always looking for interesting problems to solve. If you have a data challenge or want to collaborate, reach out!
🧠 Git-Mind Jul 2026 🔒 making public soon
A multi-agent platform that orchestrates 8 parallel LangGraph agents to comprehensively analyze GitHub repositories:
| Agent | Role |
|---|---|
| 🏗️ Architecture Analyzer | Maps repo structure, dependency graphs, tech stack |
| 🔒 Security Scanner | Runs semgrep rulesets against the codebase |
| 🦀 CVE Scanner | Rust PyO3 module for high-performance vulnerability lookup |
| 📜 Web3 Auditor | Solidity smart contract vulnerability analysis |
| 💬 RAG Chat | ChromaDB-powered Q&A over the entire codebase |
| 👀 PR Reviewer | Automated code review with context-aware suggestions |
| 🧨 Crisis Simulator | What-if scenario testing for dependency compromises |
| 📋 Report Generator | Aggregates all agent outputs into structured reports |
Key decisions: Uses Groq for low-latency LLM inference, LangGraph for agent orchestration with state persistence, and ChromaDB for vector storage enabling repository-wide semantic search. The Rust CVE scanner runs as a PyO3 module for native-speed vulnerability matching.
🛡️ crate-security-auditor Jul 2026 🔒 making public soon
Scans Rust crates for CVEs, license conflicts, and stale dependencies via the Blocks Network API.
- CVE detection — cross-references crate versions against known vulnerability databases
- License scanning — detects incompatible licenses (GPL vs MIT vs proprietary)
- Staleness checks — flags dependencies that haven't been updated in configurable thresholds
- CI-ready — exits non-zero on findings, supports fail-on severity levels
- Output formats — human-readable TUI and JSON for pipeline consumption
📦 Dep-Age Jul 2026
A CLI tool that answers one question: how old are your dependencies? Queries crates.io, npm, and PyPI registry metadata in parallel.
- Color-coded aging — 🟢 fresh, 🟡 aging, 🔴 ancient, ⚫ abandoned
- Multi-registry — supports
Cargo.toml,package.json,pyproject.toml,requirements.txt - Parallel fetches — queries registries concurrently for fast results
- CI integration —
--fail-onflag to enforce freshness thresholds in pipelines - Scriptable — JSON output for downstream tooling
- Caching — registry responses cached to avoid rate limits on repeated runs
- Downloads - 2500 + on npm
Install: cargo install dep-age or npm i -g dep-age
🔍 PipeChecker Jul 2026
A CI/CD pipeline auditor that catches issues before they hit CI — saving developer time and pipeline credits.
- Auto-detection — scans for
.github/workflows/,.gitlab-ci.yml,.circleci/config.yml - Circular dependency detection — finds cycles in job dependency graphs
- Secret scanning — flags hardcoded credentials, API keys, and tokens in workflow files
- Syntax validation — per-provider YAML/JSON validation with actionable error messages
- Interactive TUI — browse and fix issues interactively
- Watch mode — re-scans on file changes during development
- Pre-commit hook — built-in hook registration to catch issues before commit
- Downloads - 2000 + on npm
📈 Stochastic-Risk-Engine Mar 2026
A real-time quantitative risk engine that runs 10,000+ Monte Carlo simulations to model financial asset price paths.
- Monte Carlo engine — geometric Brownian motion with configurable drift, volatility, and time horizon
- Risk metrics — calculates Value at Risk (VaR) and Conditional VaR (CVaR) at multiple confidence levels
- Real-time visualization — interactive Plotly charts showing price path distributions, VaR bands, and risk contours
- Parameterizable — adjust simulation count, confidence intervals, and asset parameters at runtime
- Portfolio support — extendable to multi-asset correlation modeling