Crafting Digital Realities
Full Stack Developer & Creative Technologist building high-performance web platforms and robust digital solutions.
Explore my workCrafting visual stories through code.
I'm Roham Andarzgou, a passionate Full Stack Developer and creative professional from Iran. With years of experience in the industry, I specialize in bridging the gap between visual storytelling and technical engineering.
My passion lies in pushing the boundaries of what's possible across web and mobile applications. From building high-performance Flutter applications to scalable web solutions and code-driven motion graphics, I bring a multi-disciplinary vision to every project.
Full Stack
Technologies and tools I leverage to build scalable, high-performance web and mobile applications.
Frontend
Backend
Application
Database
MyServices
Web Design & Development
Building highly scalable, robust, and performant web applications tailored to your business needs. I focus heavily on delivering a modern, intuitive UI/UX design combined with responsive layouts that work flawlessly across all devices, backed by a secure and efficient backend architecture.
Web Design & Development
Building highly scalable, robust, and performant web applications tailored to your business needs. I focus heavily on delivering a modern, intuitive UI/UX design combined with responsive layouts that work flawlessly across all devices, backed by a secure and efficient backend architecture.
App Design & Development
App Design & Development
Python Development
Python Development
Coded Motion Graphics
Coded Motion Graphics
MyServices
My Workflow
A proven workflow that delivers exceptional results across Web and App development.
Discovery & Planning
We start by defining the roadmap—whether it's a scalable web platform or a cross-platform mobile app. I analyze your requirements to select the perfect tech stack.
How I Build
A proven process designed to transform complex workflows into scalable AI-powered systems — efficiently and strategically.
Advanced AI & Tools
Top-Tier LLMs & SDKs
- Prototyping & UX mapped in Figma and Google Stitch.
- Structured specs & plans feed top-tier AI models.
- Zero-shot correctness over naive vibe coding.
Advanced AI & Tools
Top-Tier LLMs & SDKs
- Prototyping & UX mapped in Figma and Google Stitch.
- Structured specs & plans feed top-tier AI models.
- Zero-shot correctness over naive vibe coding.
Rich Context Feeding
Spec & Knowledge Context
- Feed detailed system specs and architectural blueprints.
- Incorporate codebase-memory graphs and documentation.
- Supply high-fidelity context for zero-shot accuracy.
Rich Context Feeding
Spec & Knowledge Context
- Feed detailed system specs and architectural blueprints.
- Incorporate codebase-memory graphs and documentation.
- Supply high-fidelity context for zero-shot accuracy.
Multi-Persona Audit
Persona-Driven Auditing
- Run specialized, customized AI personas for security checks.
- Automatic SEO optimization, performance auditing, and linting.
- Double-loop feedback systems to hunt and patch bugs.
Multi-Persona Audit
Persona-Driven Auditing
- Run specialized, customized AI personas for security checks.
- Automatic SEO optimization, performance auditing, and linting.
- Double-loop feedback systems to hunt and patch bugs.
My Projects
Explore my web applications, interactive websites, developer tools, and open-source systems.Explore my web applications, interactive websites, developer tools, and open-source systems through a desktop OS window experience.

Zenth
2026Autonomous self-learning crypto paper trading terminal with multi-exchange feeds, adaptive memory, and an interactive TUI. Ingests live candlesticks across 6 exchanges, runs a quantitative SMA/RSI strategy engine under strict risk caps, and distills every losing paper trade into plain-English rules that automatically filter future setups — all recorded in a pluggable memory layer.

SubStream AI
2026A high-performance, context-aware SRT subtitle translator powered by Google's Gemini 3 Pro. Built with React, Tailwind, and client-side AI processing.

OpenClaw for Free
2026Turnkey multi-agent AI deployment framework and cloud control station running autonomous agents (OpenClaw) and secondary LLM inference VMs (llama.cpp) across free cloud VPS containers (Daytona Cloud & Freestyle.sh). Features edge relays on Railway to bypass hypervisor egress blocks, OmniRoute multi-provider LLM multiplexing, a modern mouse-driven TUI control panel, and a zero-setup Telegram management bot.

MiMo2API
2026OAuth to OpenAI-compatible API router for Xiaomi MiMo Code Desktop. Unlocks invitation-only free preview frontier models (MiMo-X-Pro-Preview & MiMo-X-Flash-Preview) across external coding agents, IDEs, and harnesses (Claude Code, Codex, Antigravity, OpenCode). Features proactive token refresh, SSE streaming with reasoning delta preservation, multi-account pooling with automated 429 failover, and dual runtimes (local CLI daemon & Cloudflare Worker).

Web2API
2026Web-to-OpenAI-compatible reverse proxy bridge for FreeModels.Pro. Unlocks web-chat frontier models (Claude Sonnet 5, Claude Fable 5, GPT 5.6 Sol/Terra, GLM 5.2, Kimi K3) for agentic coding tools like Claude Code, Cursor, Codex, and Antigravity. Features intelligent prompt coercion converting OpenAI function schemas into strict JSON directives, resilient balanced-brace JSON repair, SSE streaming with reasoning traces, and dual-runtime deployment (Hono Node daemon and Cloudflare Worker).
More Coming Soon
New web applications and developer tools in active development.

