Hi there, I'm Rahul Raj 
I'm Rahul Raj, a Full Stack developer based in Warsaw, Poland 🇵🇱, building AI-powered developer tools and security products. I founded and build the ShipIt platform — an ecosystem of AI software-engineering agents — alongside autonomous AI security tools like Sentinel and PenAgent.
What I do:
- 🏗️ Design and ship complete products end-to-end — from the database schema and API design, through agent runtimes and LLM orchestration, to polished React frontends and native Swift apps.
- 🤖 Build AI agents that do real engineering work — coding agents with repo-aware task execution, autonomous GitHub teammates that fix issues and open PRs, QA agents that drive real browsers, and pentest agents that find and prove real vulnerabilities.
- 🔐 Apply LLMs to offensive & defensive security — automated penetration testing, SAST with runtime↔code correlation, AI-assisted malware reverse-engineering, and auditor-ready reporting.
- 📊 Care about the unglamorous parts — observability, tracing, cost allocation, prompt governance, PII masking, multi-tenancy, and CI/CD.
My journey: I started with Django and React, building products across edtech, automotive, insurance, real-estate, and health domains — and over the years moved deeper into AI/ML engineering: RAG pipelines, local Llama deployments, LangChain apps, fine-tuning, and finally fully autonomous multi-agent systems. Today my daily work is designing agent loops (Plan → Act → Observe → Reflect), tool-use protocols, MCP integrations, and the infrastructure that makes agents reliable in production.
Stack in one line: Python (Django/DRF, FastAPI) · TypeScript (React 19, Node) · Swift (iOS/macOS) · Claude, OpenAI & Llama · Postgres · AWS/OVH/Docker.
I love Linux-kernel-based OSs (Ubuntu, macOS), local-first software, and open source — and I'm always exploring how far autonomous agents can go.
⚡ The ShipIt Platform — an AI software-engineering ecosystem · github.com/shipiit
| Project | What it is | Stack |
|---|---|---|
🤖 ShipIt Agent  |
Production-grade Python agent runtime — tools, MCP, hooks, skills, RAG, memory, sessions, reasoning & streaming packets. Powers a chat-based coding agent with live SSE streaming, Claude/Codex support, repo-aware task execution, code review, testing & local git automation, plus a role-based React control plane with a live Vibe-coding panel · landing page | Python, React 19 |
| ⌨️ ShipIt CLI | AI coding agent for the terminal — 65+ tools, 10+ LLM providers, computer use, browser automation, deep research, MCP, two-way Slack/Discord/Telegram/WhatsApp bots | TypeScript |
| 🔨 ShipIt Forge |
Autonomous GitHub coding agent — a teammate that fixes issues, opens PRs, reviews code with a security lens, and auto-fixes CI. Multi-provider: Vertex Gemini, Bedrock, OpenAI, Anthropic | TypeScript |
| 🧪 ShipitQA | Evidence-first autonomous product testing with real Playwright browsers, human takeover (OTP/login boundaries), and reproducible engineering reports | Python |
| ☁️ ShipIt Cloud | Cloud control plane for running ShipIt agents remotely | Python |
| 📊 ShipIt Watchtower |
Observability for LLM applications — tracing, agent graphs, prompt governance, cost allocation, PII masking. Python 3.11+, zero dependencies | Python |
| 🎨 shipit-ui-design |
Senior UI/UX in your terminal — a Claude Code plugin: 11 skills, 9 commands, 9-category design rubric, screenshot critique loops, design-system bootstrap | TypeScript |
| 🌈 ShipIt Palette |
Pick a color, ship the palette — modern color-palette studio with 14 export formats, live UI playground & full accessibility tools | JavaScript |
| Project | What it is | Stack |
|---|---|---|
| 🛡️ Sentinel | Autonomous AI penetration testing — a multi-agent platform (23 autonomous ReAct attacker agents) that finds, validates & explains real vulnerabilities in web apps and APIs, with code-level root cause, OWASP mapping, live exploit proof and auditor-ready reports · also a multi-tenant SaaS | Python, Django / DRF, React |
| 🔍 PenAgent | AI-orchestrated penetration-testing agent — fully local, autonomous Plan → Act → Observe → Reflect loop, web pentest + SAST + runtime↔code correlation, Textual TUI + web dashboard | Python |
| 🧠 AI Reverse Engineer | AI-assisted malware file analysis and reverse-engineering tooling | Python |
| Project | What it is | Stack |
|---|---|---|
| 📱 MacPilot | Securely control your Mac from your iPhone — full desktop + terminal access, built on Tailscale, SSH & Screen Sharing | Swift |
| 📸 Snappilot |
Free, open-source Snagit alternative for macOS — capture, annotate, record & OCR. Native Swift, on-device, private | Swift |
| 🔑 KeyVault |
Local-only, zero-knowledge password manager for Chromium browsers — AES-GCM-256 vault, PBKDF2 key derivation, RFC 6238 TOTP 2FA, smart autofill (works on React/Vue apps), password Watchtower, importers for 1Password/Bitwarden/LastPass. No backend, no account, no telemetry · 697 tests, 97% coverage, two runtime deps | JavaScript |
- 🧩 agent-builder — visual builder for AI agent workflows
- 📝 markdown-task-notebook — local-first markdown notebook with notes table, split editor, live preview, tags & images
- 🔔 toast-notifications — modern, customizable toast notification library for React
- 🏠 Django_RealEstate — real-estate management system with agents, listings & advanced search
- 💬 Talk to me about AI agents, LLM tooling, or application security
- 🌍 Website: iamrraj.com · Org: github.com/shipiit
- 📫 Reach me at rahul@iamrraj.com or LinkedIn