Senior Software Engineer at Wingie Enuygun Group, based in Istanbul.
Mathematical Engineering graduate (Honors) from Yıldız Technical University. Currently pursuing an MSc with a thesis on low-light image enhancement using diffusion models.
9+ years building scalable backend and platform systems across travel, vehicle finance, tourism, IoT, and SaaS. Active in open source and the Turkish developer community.
Languages: Turkish, English
- Building WingieOne, Wingie Enuygun Group’s B2B travel platform
- Developing kprompt — an open-source AI Kubernetes CLI (kprompt.ai)
- Researching diffusion models for low-light image enhancement (MSc thesis)
| Area | Technologies |
|---|---|
| Languages | C#, TypeScript, Python, Go |
| Frameworks & Tools | .NET, Node.js, React Native, Entity Framework, LangChain, Cursor |
| Infrastructure | Docker, Kubernetes, Helm, ArgoCD, Pulumi, Jenkins |
| Cloud | AWS, GCP, Azure |
| Data & Messaging | MSSQL, MySQL, PostgreSQL, MongoDB, Redis, Kafka, RabbitMQ, Firebase |
| Observability & Platform | Prometheus, Grafana, OpenTelemetry, ELK Stack, OpenShift, API Gateway, BFF |
| Architecture | Microservices, event-driven systems, CI/CD, GitOps |
Wingie Enuygun Group’s B2B platform for travel partners and agency workflows.
- Designing and delivering scalable backend services for booking, inventory, and partner integrations
- Operating high-traffic travel systems with a focus on reliability and operability
- Built SEGGY, an AI travel assistant integrating ChatGPT, LangChain, and Gemini
- Migrated monolith modules to microservices (.NET 8, Node.js)
- Introduced Kafka-based messaging; deployed on Azure and GCP with Kubernetes
- Founding engineer of a mobile analytics product (500 Istanbul–funded)
- Provisioned and deployed microservices with Pulumi and ArgoCD
- Stack: React Native, MongoDB, Firebase
- Built a cross-platform mobile app for vet consults, vaccinations, and reminders
- Pitched the product and secured investment for continued development
- Delivered credit systems and legacy migrations with C# and Entity Framework
- Moved workloads to containerized microservices with Redis, RabbitMQ, and ELK-based observability
kprompt — AI Kubernetes CLI
Natural language compiles into a reviewable plan; nothing mutates the cluster until you approve.
kprompt.ai · Examples · Discussions
| Capability | What it does |
|---|---|
| Intent compiler | English → typed plan (actions, risk, hard denies). Wipe-class prompts are refused. |
| Plan before apply | Interactive y/N or explicit --approve for mutate, Helm, and day-2 ops. |
| Observe agent | Namespace watch → incidents → gated Slack/webhook. Autopilot is propose-only; heuristic mode runs offline with zero LLM spend. |
| Day-2 integrations | Helm, Prometheus, OpenTelemetry, Grafana, and GitOps under one approval loop. Local BYOK — your kubeconfig and your LLM keys. |
| CI-ready | --output json PlanResult for automation gates. Providers: OpenAI, Anthropic, Gemini, Groq, Ollama. |
curl -fsSL https://kprompt.ai/install | bash
# brew install kprompt/tap/kprompt
kprompt "scale api to 10" # plan only
kprompt "scale api to 10" --approve
kprompt agent run -n payments --health --heuristicOffline walkthrough (kind + deliberate failures, no API key): kprompt-examples
- ISIF 2020 Gold Medal — thermal sensor–based healthtech innovation
- Patent — Contactless Temperature Monitoring System
- TÜBİTAK 2209-A research grant recipient
- Innovation Awards — Doğuş Teknoloji (2018, 2019)
Co-founder of The Coderverse — live coding on YouTube, open-source collaboration, and tech talks for Turkish developers.
“Empty your mind, be formless, shapeless, like water… Be water, my friend.”