|
Software Engineer @ Native AI

Hüseyin Höbek.

I build production AI infrastructure — from architecture to deployment.

RAG systems, data pipelines, and automation platforms. I don't just write code — I architect the system, build it, deploy it, and keep it running in production.

5+
Years of
Software
Engineering
15+ Production Systems
3 Industries Automated
End-to-End Development → Deployment

About

Building systems. Solving processes.

My focus isn't just building screens — it's understanding real problems and turning them into working, scalable systems. From RAG-based document intelligence to survey-data classification engines, from automation pipelines to cloud migrations, I've built systems that run in production.

I currently work as a Software Engineer at Native AI, shipping production AI and data infrastructure: RAG systems over meeting and document data (Supabase pgvector, OpenAI), async data-processing pipelines with FastAPI + Celery + Redis, and cloud-native deployments on AWS ECS Fargate (RDS, Secrets Manager, Cloud Map, CloudWatch) with GitHub Actions CI/CD.

What I Build

Core focus areas and the types of problems I solve.

AI & RAG Systems

Retrieval-augmented generation over documents and meetings. Vector search with Supabase pgvector, Pinecone, OpenSearch and BGE-M3 embeddings; scope-first retrieval, semantic search, and multi-intent query routing.

Automation & Workflow Engineering

Complex n8n workflows on self-hosted infrastructure. Multi-intent routing, webhook orchestration, and LLM-integrated pipelines — collection, transformation, reporting, and notifications in one system.

Data Pipelines & Transformation

Survey/SPSS/interview/review data classification and transformation. Parse-time classifiers (deterministic heuristics + calibrated logistic regression), async processing with Celery/Redis.

Cloud Infrastructure & DevOps

Production services on AWS ECS Fargate, RDS PostgreSQL, Secrets Manager, ECR, Cloud Map and CloudWatch. Docker multi-container architectures and GitHub Actions CI/CD (auto-review → staging → prod).

Backend Engineering

REST APIs with FastAPI, Celery, Redis; JWT auth boundaries, Alembic migrations, and PostgreSQL. Async worker architectures and scalable service design.

Mobile Development

User-focused mobile apps with iOS/SwiftUI and Flutter. Projects enriched with on-device ML (TensorFlow Lite), OpenAI integration, and Firebase backend.

AI Meeting & Document Intelligence Platform.

RAG system over meeting transcripts and documents. Scope-first retrieval (entity/date/type resolution before semantic search), multi-intent query routing, and an audio ingestion pipeline — built on n8n with Supabase pgvector and OpenAI.

n8n pgvector OpenAI PostgreSQL

Survey Data Processing & Classification Engine.

Parse-time dataset classifier (survey/interview/review) with a two-stage hybrid architecture: deterministic heuristics plus a calibrated logistic-regression fallback. Async DOCX/data parsing via Celery workers on a two-container ECS deployment.

FastAPI Celery scikit-learn AWS ECS

Cloud Infrastructure Migrations.

Migrated automation workloads from EC2 to ECS Fargate — headless services with Cloud Map discovery, RDS with SSL, Secrets Manager, and CloudWatch. Deployed production services from scratch with GitHub Actions CI/CD.

ECS Fargate RDS Secrets Manager CI/CD

QuickRoute — Inventory & Sales Management.

End-to-end management system that digitized retail operations, eliminated manual tracking, and provided real-time inventory-cash visibility. Unified barcode sales, customer accounts, and cash management in a single panel, built on a reporting-optimized MySQL REST API architecture.

PHP MySQL REST API JavaScript

Technical Service & Repair Tracking System.

Management system that reduced the operational complexity of service businesses into a single digital flow, making the process end-to-end traceable. Digitized the full lifecycle from device intake to delivery — including prepayments, credit tracking, service notes, cash transactions, and reporting.

PHP MySQL JavaScript State Machine

ParentAssistant — Digital Behavior Analytics.

Mobile application addressing digital parenting needs by combining device usage monitoring, NLP-based content analysis, and AI-powered reporting. Real-time sync on Firebase backend, on-device NLP model (.tflite) for fast classification, and OpenAI API for in-depth behavioral analysis.

SwiftUI Firebase TensorFlow Lite OpenAI API

Clients & Collaborations

Brands I've built software, automation and infrastructure for.

ISM Bullet Leader Munitions Tuna Metal Runbase Poseidon Machine Gazihan Lojistik Design Phili

Tech Stack

Tools and technologies in my daily workflow.

Languages
Python TypeScript JavaScript Swift Dart PHP SQL
Backend & AI
FastAPI Celery Redis OpenAI API RAG scikit-learn
Frontend & Mobile
React Next.js SwiftUI Flutter HTML CSS
RAG & Vector DB
Supabase pgvector Pinecone OpenSearch BGE-M3
Cloud & DevOps
AWS ECS Fargate RDS S3 Secrets Manager ECR Cloud Map EC2 Route 53 Docker Nginx GitHub Actions
Databases & Automation
PostgreSQL MySQL Redis Supabase Firebase n8n
Training & Technical Foundation
2021 – 2022
iOS & Swift Development

iOS app development with UIKit and SwiftUI. Architectural patterns, state management, and deployment processes.

2022 – 2023
UI/UX Design Foundations

User interface design with Figma. Applied design decisions across mobile and web projects.

2023 – 2024
Full-Cycle Web Development

Comprehensive training in frontend/backend development, database design, API architecture, security fundamentals, and AI integration.

Contact

Have a project in mind? You're in the right place.

Share your project idea, technical requirement, or collaboration proposal. I respond promptly.

Ankara, Turkey

+90 (533) 202-29-15

Message on WhatsApp

Let's discuss your project.

Your message has been sent successfully.
Something went wrong. Please try again.