Melih Sahtiyan

Backend Engineer | .NET, Node.js

About

I am a backend developer based in Istanbul, working mainly in C#/.NET 9 and TypeScript/Node.js. I graduated from Maltepe University in 2024 with a B.Sc. in Software Engineering taught fully in English, and I have spent the time since delivering freelance backends where I owned the architecture, the database schema and the deployment on my own. Two patterns run through my work. The first is architecture as a practical tool, not decoration: Clean Architecture, CQRS and Domain-Driven Design applied where they earn their keep - a Windows Service consolidating 1M+ supplier product records, a manufacturing tracker following 100K+ items through a custom workflow. The second is AI treated as engineering rather than magic: an LLM agent that extracts invoice data with LangChain and RAG, and a video generation pipeline running on a single 16 GB GPU where every scope, cost and architecture decision stayed with me.

Experience

Freelance Backend Developer

Erdag Product-Sync

Nov 2025 — Dec 2025

Istanbul, Turkiye

A B2B automotive spare-parts company needed product data from multiple suppliers consolidated into one central platform. • Designed a Windows Service with Clean Architecture and CQRS; integrated REST- and SOAP-based supplier APIs into a unified data pipeline through DDD domain modelling • Built high-performance batch processing handling 1M+ product records; upserted 140K+ active records daily using SqlBulkCopy • Implemented automated scheduling with Quartz.NET and a fault-tolerant Telegram notification system to monitor data consistency

Backend Developer Intern

Testinium

Aug 2023 — Sep 2023

Istanbul, Turkey

• Worked in an Agile team using Jira and Azure DevOps; participated in sprint planning and daily standups • Reviewed production backend codebases, focusing on debugging workflows and performance-optimization patterns

Education

B.Sc. in Software Engineering (100% English)

Maltepe University

Istanbul

2020 – 2024 · GPA: 3.14 / 4.00

High School Diploma

Doga College

2015 – 2019

Skills

AI & LLM

Claude Code · LangChain / RAG · LLM APIs (OpenAI, Claude, Gemini) · Ollama, ComfyUI, SDXL · PyTorch, Weights & Biases

Architecture & Patterns

Clean Architecture · CQRS · Domain-Driven Design · Repository Pattern · Queue-driven worker systems · REST API design · SOAP integration

Currently learning

Next.js · Supabase · Vercel · FastAPI · MCP

Data & Infrastructure

PostgreSQL · MS SQL Server · MongoDB · RabbitMQ · Docker / Docker Compose · AWS (EC2, ECS, RDS) · Quartz.NET

Languages & Frameworks

C# / .NET 9 · TypeScript / Node.js · Git / GitHub · Python · React

Testing

Jest / Mocha

Languages

TurkishNative
EnglishB2 - Professional Working Proficiency
GermanA1 - Beginner

Projects

Content Creation AutomationFeatured

An autonomous pipeline that turns a single video idea into a finished 9:16 vertical short: LLM scenario breakdown, identity-locked per-scene frames, animation, audio, assembly and upload. It runs on a single 16 GB GPU, and that constraint shaped most of the design.

NestJS
RabbitMQ
PostgreSQL/pgvector
LangChain
ComfyUI
SDXL
FFmpeg

VAT Reconciliation ToolFeatured

Automated VAT reconciliation against ledger records, on a backend built with CQRS and Domain-Driven Design. I built an LLM agent with LangChain and RAG that extracts data from invoices, detects inconsistencies and suggests corrections.

C#/.NET
Node.js
LangChain.js
RAG
CQRS
DDD

Social Media Backend

Backend handling 10K+ requests at sub-150ms latency, with data integrity ensured by 150+ unit tests and 60%+ code coverage. The technical articles on clean code and the anemic domain model came out of this project.

Node.js
Express.js
Mocha
Jest

Source

LangChain Agent

LLM agent built on NestJS.

NestJS
LangChain

Source

All 7 projects →