Canberk Karaerkek
Canberk Karaerkek

AI / Agentic Systems Engineer & Architect

Open to offers · Member since 27 Aug 2026
Message
Location
Ankara, Türkiye
Desired salary
Unspecified
Work preference
Remote Only / Full Time, Part Time, Contract, Freelance, Temporary
Experience level
Lead

About

Professional summary

I am an AI and agentic-systems engineer and architect specializing in trustworthy LLM-based systems, multi-agent orchestration, AI safety, and local-first AI infrastructure.

I design autonomous AI agents that are auditable, governed, and safe by construction. My work focuses on proposal-bound actions, execution gating, provenance tracking, immutable audit trails, hallucination guards, and robust operational controls.

I am the founder of the DoppleGround foundation and creator of NEXUS, a local-first, post-quantum-immune agentic operating system. Through NEXUS, I build multi-agent workflows, A2A governance mechanisms, on-device inference capabilities, and agent-driven visual AI pipelines.

I work across LLM engineering, generative AI, edge ML, WebGPU compute, GGUF quantization, Ollama, and post-quantum cryptography. I have published more than 140 open-source repositories spanning agents, kernels, datasets, and applied LLM systems.

My background in applied and engineering physics gives me a first-principles approach to AI engineering. I bring experience in semiconductor devices, nanotechnology, thin-film deposition, scientific programming, simulation, and experimental research to the development of reliable AI systems.

I am open to roles and collaborations in agentic AI, AI safety, LLM systems, and local-first AI infrastructure.

Notice period: 3 days

Skills

12 capabilities

Tech stack & tools

Working toolkit

Application Hosting

Data Stores

Design

Experience

Career history

Founder & AI Systems Architect · Curator of Machine Cognition DoppleGround foundation

I founded the DoppleGround foundation and architect NEXUS, a local-first, post-quantum-immune agentic operating system for governed and auditable AI.

I build multi-agent infrastructure with A2A governance, proposal-bound and execution-gated actions, provenance tracking, and immutable audit ledgers. My work also covers adversarial robustness, alignment-failure mitigation, hallucination detection, post-quantum cryptography, Ollama-based local inference, GGUF quantization, WebGPU kernels, LiteRT/TFLite edge runtimes, and agent-driven visual generation pipelines.

I maintain more than 140 public open-source repositories covering agents, kernels, datasets, and applied LLM systems. I apply first-principles physics, including Landau-Ginzburg phase-transition theory, to model and detect LLM hallucinations.

Physical-Technical Assistant/Engineer beeOLED

Independently operated machines used to produce organic semiconductor devices and completed front-end preparation activities.

Performed post-processing component characterization, documentation, and photophysical characterization of emitter materials for suitable device structures. Supported laboratory maintenance and improvements to production throughput and quality.

Part-time Project Work Fraunhofer-Gesellschaft

Contributed to a reactive magnetron thin-film coating project focused on optimizing complex compound layers.

Applied nanoparticle-source sputtering techniques to coat nanomaterials with compound layers and collaborated with multidisciplinary teams to integrate processes and technologies.

Master Thesis Student Fraunhofer-Gesellschaft

Developed a novel reactive magnetron sputtering PVD component and improved nanocomposite-layer production efficiency. Led TiO2 thin-film coating optimization through precise process control and monitoring.

Produced sustainable TiO2 layers using a PID-controlled reactive coating process and authored a comprehensive master's thesis. Used scientific and optical acquisition tools including OED, collimator design, fiber-optic high-vacuum arrangements, ellipsometry, profilometry, and UV-Vis RT.

Working Student Helmholtz-Zentrum Dresden-Rossendorf (HZDR)

Used scientific programming and data visualization to support materials research for liquid-metal battery technology.

Developed mathematical models and simulations to analyze elemental capabilities and support battery-performance optimization.

Education

Learning history

Technische Universität Dresden

Master of Science, Organic and Molecular Electronics

Master of Science studies in Organic and Molecular Electronics.

Hacettepe Üniversitesi

Bachelor of Science, Engineering Physics

Bachelor of Science studies in Engineering Physics.

This professional hasn’t added portfolio projects yet.

This professional hasn’t listed any services yet.

People also viewed

All talent ›
Jobs Talent AI Tools Salaries
Menu