Mustafa

AI engineering, full-stack product systems, and typed developer tools.

I build fast web and mobile applications, experiment with agentic software, and keep reaching for tools that make systems simpler, safer, and easier to evolve.

What I am focused on

• AI engineering that ships: agents, RAG workflows, MCP servers, voice/chat interfaces, and evaluation-minded product loops.
• Full-stack TypeScript apps with Next.js, Hono, oRPC, Drizzle, PostgreSQL, Sentry, and clean API boundaries.
• Mobile products with React Native and Expo, especially tooling, navigation, performance, notifications, and polished app structure.
• Rust as the language I study for systems thinking: ownership, concurrency, web servers, embeddings, macros, and reliable backend foundations.

My working stack

Layer	Tools I reach for
Product UI	React, Next.js, Tailwind CSS, shadcn/ui, React Native, Expo
APIs	Hono, oRPC, REST, GraphQL, MCP servers
Data	PostgreSQL, Drizzle ORM, Prisma, SQL-first schema design
AI systems	LangGraph, RAG, agent workflows, tool calling, voice/chat UX
Quality	TypeScript, Vitest, Playwright-style browser testing, Sentry, observability
Systems learning	Rust, Rocket, concurrency, macros, embeddings, memory management

Public work signals

These repositories show the areas I keep returning to:

• langgraph-workflows: LangGraph.js workflows and AI orchestration practice.
• hono-sequential-mcp-server: TypeScript backend experiments around Hono and MCP.
• hono_with_vite: Hono with a modern Vite setup.
• orpc-contract-example: contract-first TypeScript APIs with oRPC.
• simple-mcp: MCP learning and server-side integration experiments.
• rigel-rn: React Native starter architecture with Redux, sagas, maps, notifications, hooks, and performance topics.
• rigel-cli: React Native CLI tooling for screens, components, Redux files, and app structure.
• rust-embeddings: Rust embeddings exploration with fastembed and ONNX Runtime.
• rust-web-server-example, rocket-101, and concurrency: Rust backend and systems practice.

How I think about software

I like systems that are boring in production and sharp in development: typed contracts, small composable services, clear database models, observable failures, and UIs that make complex workflows feel direct.

My current direction is the intersection of AI-native product engineering and practical full-stack architecture: giving agents real tools, real state, real feedback loops, and interfaces people can trust.

GitHub activity

GitHub's native contribution graph is the source of truth for my activity. It is a GitHub-counted view, not a raw total of every commit or PR across every branch, fork, or private workspace.

GitHub contribution metric	Count
GitHub-counted contributions	3,264
Public commits	379
Public pull requests opened	64
Public pull request reviews	0
Public issues opened	1
Public repositories created	93
Restricted/private contributions	2,727

Last synced from GitHub: 2026-10-21.

Recent public activity

• 2026-10-20: Merged pull request #137 in rel-kit/relkit.
• 2026-10-20: Pushed to rel-kit/relkit on fix/release-publishing-window.
• 2026-10-20: Opened pull request #137 in rel-kit/relkit.
• 2026-10-18: Pushed to rel-kit/relkit on fix/application-database-ownership.
• 2026-10-20: Merged pull request #135 in rel-kit/relkit.

GitHub exposes restricted/private contributions as a total, not as a commits vs PRs breakdown. The commit and PR rows above are the public breakdown GitHub exposes, not lifetime totals.