Skip to content
View Mehdi-HS's full-sized avatar
💭
:|
💭
:|

Block or report Mehdi-HS

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Mehdi-HS/README.md
Mehdi Hoseini — Full-stack developer working across AI and Linux

I work across product, software, and infrastructure—and I like seeing the whole thing through.



English  ·  فارسی  ·  Portfolio  ·  Email

01 / HELLO

Hi, I’m Mehdi.

I’ve worked in software since 2017. I started with interfaces and frontend work, then gradually moved deeper into backend systems, AI, Linux, and infrastructure. What keeps me interested isn’t code on its own; it’s understanding an unclear problem, finding a sensible way through it, and staying with the work until the result is useful and maintainable.

I’m a co-founder of Ravand Studio, where I lead engineering and AI. Depending on what a project needs, I may spend one day on the product experience and the next on an API, an LLM workflow, a deployment, or a server issue. I enjoy moving between those layers because it keeps technical decisions grounded in the product.

I value clear teamwork, small releases we can learn from, practical documentation, and systems that remain dependable outside a demo. I’d rather discuss uncertainty and constraints early than pay for vague decisions later.

Much of my work is developed in private GitLab and self-hosted repositories, so my public GitHub activity shows only part of my experience.

02 / LATELY

What’s on my desk
Current focus Building AI products and the operational tools behind them at Ravand Studio and HoosheMasnoei
In the details Service reliability, model cost, failure paths, payments, and support workflows
Open question How to keep a product simple without trading away maintainability

03 / A FEW THINGS I’VE BUILT

The digital product studio I co-founded. I lead engineering and AI and stay involved from defining the problem and architecture through release and maintenance.

STUDIO ENGINEERING AI

A Persian platform for accessing text, image, and video models. Alongside the model integrations, I work on subscriptions, payments, support flows, and the internal tools needed to operate the product.

AI PRODUCT FULL-STACK OPERATIONS

A nutrition and wellness product that uses meal photos, progress tracking, and AI-assisted guidance to offer a simple, practical experience.

AI HEALTH PRODUCT

A financial platform developed in a private, self-hosted GitLab environment. My work has focused on the frontend and the build and deployment process.

FINTECH FRONTEND CI/CD

An AI workspace built from connected frontend and backend services and deployed on private infrastructure.

AI WORKSPACE FULL-STACK

My longer-form portfolio: selected work, what I can help with, and a more complete picture of how I approach a build.

PORTFOLIO NOTES

04 / THE TOOLBOX

I’m not attached to a stack for the sake of it. These are simply the tools I know well enough to trust when the fit is right:

  • Interfaces: TypeScript, React, Next.js, Tailwind CSS, React Native
  • Services & data: Node.js, NestJS, Laravel, MySQL, Redis
  • Systems: Linux, Docker, Kubernetes, Terraform, GitLab CI
  • Product work: Figma, interface architecture, design systems, release planning
How I like to work
  • Start with the real problem, not the most interesting technology.
  • Make the first release small enough to learn from.
  • Treat AI as part of the product, with limits and fallbacks—not as decoration.
  • Leave useful logs, readable docs, and a way back when a deployment goes wrong.
  • Be open about what we do not know yet, then make a plan to find the answer.

سلام، من مهدی‌ام.

از سال ۱۳۹۶ در حوزه نرم‌افزار کار می‌کنم. مسیر کاری‌ام از طراحی و توسعه رابط کاربری شروع شد و به‌مرور به بک‌اند، هوش مصنوعی، لینوکس و زیرساخت رسید. چیزی که برایم جذاب است فقط نوشتن کد نیست؛ دوست دارم یک مسئله مبهم را درست بفهمم، برایش راه‌حل مناسبی پیدا کنم و آن را تا جایی پیش ببرم که در دنیای واقعی قابل استفاده و قابل نگه‌داری باشد.

هم‌بنیان‌گذار استودیو روند هستم و مسئولیت مهندسی و هوش مصنوعی را بر عهده دارم. بسته به نیاز پروژه، ممکن است یک روز روی تجربه کاربری کار کنم و روز بعد درگیر API، جریان‌های LLM، استقرار یا عیب‌یابی سرور باشم. این رفت‌وآمد میان محصول و زیرساخت را دوست دارم؛ چون کمک می‌کند تصمیم‌های فنی از نیاز واقعی محصول جدا نشوند.

به همکاری شفاف، نسخه‌های اولیه کوچک و قابل‌آزمایش، مستندات کاربردی و سیستم‌هایی که در شرایط واقعی قابل اتکا باشند اهمیت می‌دهم. ترجیح می‌دهم از همان ابتدا درباره ابهام‌ها و محدودیت‌ها حرف بزنیم تا اینکه بعداً هزینه تصمیم‌های مبهم را بپردازیم.

بخش قابل‌توجهی از پروژه‌هایم در گیت‌لب خصوصی و مخازن self-hosted توسعه داده می‌شوند؛ بنابراین فعالیت عمومی گیت‌هاب فقط بخشی از سابقه کاری من را نشان می‌دهد.

این روزها

تمرکز فعلیتوسعه محصولات هوش مصنوعی و ابزارهای عملیاتی مرتبط با آن‌ها در استودیو روند و هوش مصنوعی دات‌کام
جزئیات مهمپایداری سرویس‌ها، مدیریت هزینه مدل‌ها، طراحی مسیرهای جایگزین هنگام خطا، پرداخت و فرایندهای پشتیبانی
پرسش این روزهاچطور می‌توان محصول را ساده نگه داشت، بدون اینکه کیفیت فنی و قابلیت نگه‌داری آن قربانی شود؟

چند کاری که ساخته‌ام

  • استودیو روند — استودیوی طراحی و توسعه محصول دیجیتال که هم‌بنیان‌گذارش هستم. مسئولیت مهندسی و هوش مصنوعی را بر عهده دارم و از مرحله تعریف مسئله و معماری تا انتشار و نگه‌داری محصول درگیر کار می‌مانم.
  • هوش مصنوعی دات‌کام — پلتفرمی فارسی برای دسترسی به مدل‌های متنی، تصویری و ویدیویی. علاوه بر اتصال به مدل‌ها، روی بخش‌هایی مثل اشتراک، پرداخت، پشتیبانی و ابزارهای مدیریتی هم کار کرده‌ام.
  • FitAI — محصولی در حوزه تغذیه و سلامت که با ثبت تصویری وعده‌ها، پیگیری روند و راهنمایی هوشمند، تلاش می‌کند تجربه‌ای ساده و کاربردی ارائه دهد.
  • TheFinance — پلتفرمی مالی که در بستر گیت‌لب خصوصی و self-hosted توسعه داده می‌شود. تمرکز من در این پروژه روی فرانت‌اند و فرایند ساخت و استقرار بوده است.
  • iChatGPT.dev — یک فضای کاری هوش مصنوعی با سرویس‌های فرانت‌اند و بک‌اند که روی زیرساخت خصوصی اجرا می‌شوند.

شیوه‌ای که دوست دارم کار کنم

  • پیش از انتخاب تکنولوژی، مطمئن شویم مسئله را درست فهمیده‌ایم.
  • نسخه اول را کوچک، قابل‌آزمایش و مناسب یادگیری نگه داریم.
  • برای قابلیت‌های هوش مصنوعی از ابتدا محدودیت، هزینه، خطا و مسیر جایگزین در نظر بگیریم.
  • لاگ‌ها، مستندات و فرایند بازگشت از انتشار ناموفق را بخشی از محصول بدانیم، نه کاری اضافه.
  • اگر چیزی را هنوز نمی‌دانیم، همان ابتدا شفاف بگوییم و برای پیدا کردن پاسخ برنامه داشته باشیم.

اگر چیزی برای ساختن دارید

اگر برای محصولتان به کسی نیاز دارید که بتواند میان رابط کاربری، بک‌اند، هوش مصنوعی و زیرساخت ارتباط برقرار کند، خوشحال می‌شوم درباره پروژه صحبت کنیم. برای شروع، یک توضیح کوتاه از مسئله و هدف محصول کافی است.

ایمیل · تلگرام · پورتفولیوی فارسی


05 / PUBLIC ACTIVITY

Mehdi's public GitHub contribution graph

This graph does not include private GitLab and self-hosted work.

06 / SAY HELLO

If you need someone who can connect the product experience with the systems behind it, I’d be glad to hear what you’re working on. A short note about the problem and the goal is enough to start.

Email  ·  Telegram  ·  Portfolio  ·  Ravand Studio



Karaj, Iran · building since 2017

Popular repositories Loading

  1. Black-Support-Bot Black-Support-Bot Public

    Lua 13 16

  2. blackplus_test blackplus_test Public

    Lua 4 8

  3. BlackPlus BlackPlus Public

    A new bot for manage your SuperGroups...

    Lua 1

  4. overhang.js overhang.js Public

    Forked from paulkr/overhang.js

    🔔 A jQuery plugin for notifications, prompts and confirmations.

    HTML

  5. SenjedBot SenjedBot Public

    PHP

  6. mehdi.github.io mehdi.github.io Public