Distributed systems
Microservices, event-driven workflows and messaging architectures designed for failure — retries, idempotency and back-pressure built in.
CTO @ Fleksup / Istanbul
Engineering leader and hands-on architect with 16+ years building distributed systems, event-driven microservices and cloud-native SaaS — across e-commerce, ERP, marketplace, cargo and payment ecosystems.
I lead the architecture and development of multi-tenant, event-driven microservice platforms that keep ERP systems, marketplaces, cargo providers, payment systems and financial services in sync — at high volume, around the clock.
My focus is resilient, production-grade systems: scalability, maintainability and performance by design, backed by observability you can actually trust. I stay hands-on in system design and critical implementations while defining technical strategy, platform roadmap and engineering standards for the teams around me.
A simplified view of the kind of system I design: a multi-tenant integration core, an event backbone, and dozens of external ecosystems kept in sync. Hover a layer.
Microservices, event-driven workflows and messaging architectures designed for failure — retries, idempotency and back-pressure built in.
Kubernetes on Azure AKS, CI/CD and infrastructure governance for a 50+ service ecosystem serving 1,000+ customer environments.
Prometheus, Grafana, Loki and Jaeger — metrics, logs and distributed traces wired in from day one.
Technical strategy, roadmap and governance; Agile delivery, hiring, onboarding and mentoring across frontend, backend and mobile teams.
Three engagements that shaped how I lead: one platform transformation, one integration problem at volume, one engineering organization.
Multi-platform integrator SaaS. Architecting and leading multi-tenant, event-driven platforms behind large-scale e-commerce integration operations.
Led frontend, backend and mobile teams building supply-chain optimization and distributed planning products.
ICRON's industrial engineers develop AI and optimization models for supply chains. I engineered the products they — and their customers — work in.
Enterprise web and mobility solutions for municipal and enterprise customers in the Netherlands, delivered by teams across countries.
Nine years of connecting commerce — the e-commerce and ERP integration foundation my forecasting research builds on today.
E-commerce and ERP integrations end to end; B2B, B2C and marketplace solutions; virtual POS with major Turkish banks, iyzico, PayU and PayPal.
B2B, B2C and private-shopping platforms; ERP and Microsoft Dynamics CRM integrations; consultancy for 100+ customer projects.
Full role-by-role history on LinkedIn ↗
For 16 years I've built the systems that move commerce data. Now I study what that data can predict — and how to make those predictions trustworthy enough to act on.
Products don't share one demand pattern: some sell smoothly, others are intermittent or lumpy. My research asks when a model per demand regime beats one global model — comparing global and cluster-specific stacking ensembles, and choosing between them per regime using only honest cross-validation forecasts, never the holdout.
→ A journal article based on this work is in preparation.
Retrieval-augmented generation that grounds LLM answers in real domain and platform knowledge.
Bringing ML into an integration platform — where high-volume commerce data already flows.
Taking my thesis research to production: SKU-level forecasts and smarter stock decisions for merchants.
Bilişim Yıldızları — e-Dönüşüm Competition
Open to research collaborations, talks and conversations on platform architecture, applied machine learning and demand forecasting.