Portfolio & Expertise Showcase
Explore the technologies, projects, and ideas that drive my work
Core technologies I work with
Integrated technology stacks for different applications
Languages: Go + Python
Framework: Go backend, FastAPI for ML
High-performance bidding optimization with custom ML models and hybrid architecture
Languages: TypeScript
Framework: Next.js + MUI DataGrid
Real-time analytics dashboard with fraud detection and campaign management
Languages: TypeScript
Framework: Next.js with MDX
Personal blog with MDX support, custom gallery components, and responsive design
Languages: Ruby + Elixir + Node.js
Framework: Rails, Phoenix/OTP, and Node/Fastify with a shared React/Redux front end
Multi-tenant advertising reporting dashboard serving roughly 12,000 advertisers, evolved across three successive backends
Languages: Elixir
Framework: Phoenix with LiveView
Concurrent, fault-tolerant applications with real-time capabilities
Languages: Bash / NodeJS / DHT22
Framework: Custom scripts with data logging
Hourly temperature and humidity logging using Raspberry Pi and DHT22 sensor
Work I'm proud of
Comprehensive dashboard featuring fraud detection systems, ML model management interfaces, campaign management tools, and real-time analytics. Built with TypeScript and MUI DataGrid for performance-optimized data visualization.
TECHNOLOGIES
Production Phoenix/LiveView application serving this consulting portfolio and contact pipeline, deployed on Fly.io with PostgreSQL. Includes an Nx/Axon-backed semantic project search built directly into the site.
TECHNOLOGIES
Owned the React/Redux dashboard layer of LocaliQ's multi-tenant advertising reporting platform, serving roughly 12,000 advertisers across North America, Europe, Latin America, and Asia-Pacific. Worked across three successive backends for the same product - the original Rails application, an Elixir/Phoenix OTP umbrella, and the current Node/Fastify service. Delivered a multi-tenant benchmarking sandbox, multi-location reporting access control, and geofencing/walk-in attribution reporting, and cut typical dashboard response times from roughly five seconds to under one.
TECHNOLOGIES
ASP.NET MVC with Angular 10 framework.
TECHNOLOGIES
Integration of IoT weather station data with Strava for fitness tracking and analysis.
TECHNOLOGIES
Elixir/Phoenix application with LiveView for real-time data synchronization and presence tracking. Demonstrates fault-tolerant concurrent architecture.
TECHNOLOGIES
Next.js-based blog with MDX support, custom photo gallery components, and responsive design. Documents travel experiences and technical musings with high-quality image galleries.
TECHNOLOGIES
Solo-built campaign bidding and fraud detection platform for digital advertising, in active development. Transitioned bid optimization from costly OpenAI API calls to a custom ML model, cutting cost by 70% while maintaining accuracy. Architected end-to-end - from data model through production deployment - with AI-assisted development.
TECHNOLOGIES
Devices and infrastructure
Desktop
AMD 9950X3D 4.30GHz 16-Core, 128GB RAM, 3x 2TB NVMe - 36 TB combined - custom workstation
Desktop
AMD Ryzen 5 5600X 4.65GHz 6-Core, 64GB RAM, 2 x1TB NVMe- 8 TB combined - dev machine
Laptop
Apple Silicon M4, 32GB RAM, 1TB SSD - Primary mobile station
Laptop
Apple M2, 32GB RAM, 1TB SSD - Secondary mobile station
Desktop
Apple Silicon M2, 16GB RAM, 1TB SSD - Mobile Development
Desktop
Intel i7, 32GB RAM, 512GB SSD - Web development
Laptop - Ubuntu 24.04 LTS
Intel i7 1260P, 32GB RAM, 1TB SSD - On-the-go computing
Tablet
12.9" M2 chip, Apple Pencil - Documentation and design work
Server
Intel Xeon, 32 RAM, 1TB NVMe - Development and testing environment
Laptop
Intel i5-6300U, 8 GB RAM, 512 GB NVMe - Legacy support and testing
Server
Intel Xeon, 32 RAM, 1TB NVMe - Development and testing environment
Server
Intel Xeon, 64 RAM,
NAS
AMD Ryzen R1600 8GB RAM, 2.6 GHz
NAS
Alpine AL212 1GB RAM, 1.4 GHz
Single Board Computer
16GB RAM, SSD - IoT and edge computing testing
Articles, ideas, and technical insights
Lessons learned from migrating from OpenAI API to custom ML models. Achieved 70% cost reduction through careful architecture design.
Exploring fault tolerance and concurrency patterns in Elixir. How to build systems that can handle thousands of concurrent connections.
Reflections on working remotely from a small coastal town in Turkey. Balance between exploration and productivity.
Designing maintainable microservices. Communication patterns, testing strategies, and deployment considerations.
Benefits of using TypeScript across the entire stack. Type safety, developer experience, and long-term maintenance.