Multi-cluster Kubernetes & microservices
Contribute to distributed AI infrastructure across cloud clusters, on-prem K3s, and edge deployments, connecting services across languages and messaging systems.
KUBERNETES · MICROSERVICES · VIDEO
Distributed systems.
Real-time video at scale.
I’m Burak Hamuryen, a senior software engineer in Berlin with 14+ years of experience. I build distributed systems, microservices, and real-time video streaming and processing platforms.
My work spans multi-cluster Kubernetes architectures and services written in Go, Kotlin, Scala, Python, and Java. I connect backend services, messaging, and video pipelines across cloud, on-premise, and edge environments, with a focus on performance and reliability in production.
At intenseye, I contribute to backend services, AI video pipelines, and infrastructure operating across roughly 10,000 Kubernetes pods. My experience also includes sensor and device integration, industrial I/O, and NVIDIA Orin edge systems that connect software decisions to physical equipment.
High-performance video pipelines
Build and optimize real-time video pipelines for streaming, decoding, processing, encoding, and recording. Experience spans low-latency delivery, concurrent camera streams, and AI video processing across cloud and edge environments.
Camera, sensor & device integration
Connected cameras, radar, access-control systems, and industrial I/O through device discovery, real-time data processing, and SCADA integration. Work includes a shared platform for 5,000+ cameras and sensors across roughly 50 locations.
Concurrent, transactional backends
Designed a national auction platform with live balance checks, bank integrations, and 1,000+ concurrent bids per auction.
Legacy systems back under control
Resolved memory leaks and stabilized a critical platform within one week, then helped move it to a microservice architecture.
THE JOURNEY
Professional experience.
2022 — Now
Senior Software Engineer · intenseye
Part of the core engineering team at intenseye, an AI workplace safety platform deployed across 25+ countries. Contributing across multiple languages and services in a polyglot stack — backend, AI pipelines, and infrastructure. Working across ~10,000 Kubernetes pods on Google Cloud, on-prem K3s deployments, and NVIDIA Orin edge devices shipped to customers. Industrial integrations enable real-time production line halts when safety threats are detected.
2020 — 2022
Lead Software Architect · iSIM Platform
Transitioned from Proline in 2020 when its iSIM product became a separate company, continuing work on the platform. Led architecture of a hybrid on-prem/cloud IoT orchestration platform deployed across ~50 locations. Unified 5,000+ cameras and sensors from multiple vendors into a single monitoring system with SCADA integration, big data analytics, and predictive maintenance. Built custom RTSP-to-WebRTC bridge by compiling Chrome engine for real-time 4x4 multi-camera grid rendering in browser. Central command center for live viewing, recording, video analytics, and traffic violation operations.
2020
Freelance · Ministry of Trade (Turkey)
Designed and built the national e-Auction platform, including a real-time bidding engine processing 1,000+ concurrent bids per auction with live balance checks and margin transactions. The platform has 1,430,066 active users as of September 14, 2026. Integrated with 4 major banks, e-Devlet, MERSIS, and MERNIS.
2016 — 2020
R&D Engineer · Proline Integrated Intelligence
Joined during a challenging phase of a long-running government project with critical stability issues. Resolved memory leaks and stabilized the system within one week, then redesigned the entire platform from scratch with microservice architecture and REST APIs in under a year. Scaled to 60+ services on K8s. Built centralized city management platform for an international deployment. Served as Scrum Master driving microservice adoption company-wide.
Jan 2016 — Aug 2016
Senior R&D Engineer · Ekin Technology
Solved NAT traversal for peer-to-peer video transmission using UPnP and hole punching. Built a TCP/HTTP proxy with C++ and Boost.Asio, implemented device discovery and configuration, and worked on FFmpeg video encoding and decoding.
Sep 2012 — Jan 2016
R&D Engineer · InfoDif (Information Diffusion Inc.)
Developed a native C++/Qt video management system with live viewing, recording, and analytics. Built camera and NVR administration tools, real-time radar and access-control data processing, and radar overlays synchronized with PTZ cameras. Worked on Capra image-processing pipelines and video transcoding for low-bandwidth systems.
SELECTED ENGINEERING WORK
Projects you
can explore.
Professional platforms I’ve helped build, followed by selected public repositories. From national auctions to video analytics, device integration, and cloud infrastructure.
e-İhale (opens in a new tab)
Turkey’s Ministry of Trade auction platform, with 1,430,066 active users as of September 14, 2026. Supports live bidding and integrations with banks and government registries.
My contribution Designed and developed the .NET Core platform, including real-time bidding, balance checks, margin transactions, and integrations with four banks, e-Devlet, MERSIS, and MERNIS.
I-Bex (opens in a new tab)
An InfoDif product: a multi-camera video analytics and management platform built on the Capra image-processing framework.
My contribution Developed native C++/Qt video management components for live viewing, recording, and analytics, alongside camera/NVR configuration and sensor, alarm, and notification handling.
iSIM Platform (opens in a new tab)
Originally a Proline product, iSIM became a separate company in 2020. I continued with the platform through that transition, bringing cameras, sensors, and security systems into a shared monitoring environment.
My contribution Led platform architecture for deployments across roughly 50 locations, unifying 5,000+ cameras and sensors. Built a native RTSP-to-WebRTC bridge for browser viewing and worked on SCADA integration.
intenseye (opens in a new tab)
An AI video analytics platform connecting workplace risk detection with real-time alerts and industrial device integrations.
My contribution Contribute across backend services, video and AI pipelines, and infrastructure spanning Google Cloud, on-prem K3s, and NVIDIA Orin edge devices. Work includes customer I/O integration and the systems connecting detection to device actions.
RTSPProxy.NET (opens in a new tab)
A .NET RTSP proxy, connecting native video-streaming components with the .NET ecosystem.
VideoStreamer (opens in a new tab)
A C++ RTSP video streamer built with live555. A look at the native streaming side of my work.
EArsiv (opens in a new tab)
A Java library for Turkey’s e-Arşiv / e-Fatura portal: invoice creation, queries, cancellation, SMS approval, and PDF retrieval.
MY TOOLBOX
Technical skills.
Languages
Backend & Frameworks
Cloud & DevOps
Data & Messaging
Video & AI
Domains
Devices & Industrial Systems
Edge & On-Premise
FOUNDATIONS
Education & credentials.
2020
Patent — Video Network Gateway System
Turkish Patent Office (2020/17235)
2013 — 2019
M.Sc. in Information Systems
Hacettepe University, Ankara
2008 — 2012
B.Sc. in Electrical & Electronics Engineering
Niğde University — Honours Degree
NOTES FROM THE FIELD
Writing about
what I build.
Engineering articles by Burak Hamuryen on distributed systems, AI agents, Kubernetes, real-time video, and the trade-offs behind production software.
Your Kafka Consumers Are Becoming a Workflow Engine
Read story ↗ (opens in a new tab)
When Is Self-Hosting AI Actually Cheaper?
Read story ↗ (opens in a new tab)
Python 3.15 Lazy Imports: Where the Startup Cost Goes
Read story ↗ (opens in a new tab)
Your Camera Is Online. Is the Video Still Live?
Read story ↗ (opens in a new tab)
4 Years, 6 Languages, 1 Monorepo: What I’d Do Differently
Read story ↗ (opens in a new tab)
Before Buying More GPUs, Fix Your Video Pipeline
Read story ↗ (opens in a new tab)Start with a problem
Designing reliable AI agents: retries, duplicate actions, and what to do when a tool succeeds but its response is lost.
Diagnosing a slow video pipeline: trace where frames wait and cross between CPU and GPU memory before adding capacity.
Deciding when to split a monolith: a first-hand account of service boundaries, deployment constraints, and the costs of a small team operating separate services.