HYHasan Yiğitresearch / software / systems
TRENDE

02 / SELECTED WORK

Five axes.
One engineering language.

A selected portfolio across image processing, computational electromagnetics, AI algorithms, secure infrastructure and Docker-based software infrastructure, translating academic precision into product-ready systems and compact use cases.

01 / QUEVATECHdeep-tech
Abstract secure entropy and key generation visual QuevaTech

Secure infrastructure, entropy and verifiability

Research and productisation around physical entropy, AI-hybrid TRNG, key/password management, secure communication and public services that prove the immutability of a process or software record.

ProblemRandomness, key management and process verification often live in separate, weakly connected layers.
ApproachCombine physical entropy, AI-assisted quality checks and public-proof thinking in one secure infrastructure line.
OutputA productisable frame for TRNG, password/key management and verifiable record services under QuevaTech.
TRNGsecure infrastructurepassword / key managementpublic verification
02 / IMAGE PROCESSINGresearch
Abstract image processing research visual across health, retail and other domainsvision · health · retail

Image processing systems

Computer vision research across health imaging, oral/radiological analysis, retail customer tracking and domain-adaptable visual intelligence.

ProblemHealth, radiology and retail environments require different data quality, accuracy and explainability trade-offs.
ApproachDomain-adaptable image-processing workflows, deep learning and measurable experimental setups.
OutputA reusable computer-vision practice for health, oral/radiological analysis and in-store customer/interest measurement.
computer visionmedical imagingretail analyticsdeep learning
03 / EM
Computational electromagnetics and high-frequency simulation visualEM · FDTD
Computational electromagnetics

Numerical design of high-frequency systems through FDTD, field propagation, parameter sweeps and computational engineering.

ProblemTrial-and-error design is slow and expensive.
ApproachUse FDTD, parameter sweeps and model-based optimisation to make the design space computable.
OutputA computational design workflow suitable for papers, prototypes and engineering decisions.
FDTDEM simulationcomputational electromagneticsnumerical modelling
04 / AI
AI, algorithms and computational engineering

A research practice that combines deep learning, optimisation, data processing, simulation and prototyping in one engineering workflow.

ProblemAI, simulation and software often progress as separate islands.
ApproachConnect algorithm design, experiment design and prototyping in one measurable workflow.
OutputResearch-software practice adaptable to image processing, EM and infrastructure projects.
algorithmsoptimizationsimulationresearch software
05 / INFRA
Docker-based software infrastructure

Docker/Kubernetes ecosystem work, Linux server operations, reverse proxy, load balancing, secure deployment and research operations for moving research prototypes into runnable services.

ProblemResearch prototypes often fail to become reliable, maintainable services.
ApproachBring containers, servers, proxies, load balancing and security into the design from the start.
OutputInfrastructure discipline for demos, research systems and pre-product services.
DockerKubernetesLinux serversload balancingreverse proxysecure deploymentresearch operations

APPLIED R&D

06 / TEYDEB 15052023

Computer vision in retail

Computer vision research for customer tracking and interest measurement in retail environments.

ProblemIn-store behaviour and interest measurement often remain manual or low-resolution.
ApproachUse image processing and computer vision to produce observable metrics.
OutputSector-adapted R&D and prototype-development experience.
07 / TÜBİTAK2019—21

Cyber-physical systems

Model-based design and testing projects; second place in the TÜBİTAK 2242 programme.

ProblemPhysical system design needs model, prototype and test flows to work together.
ApproachModel-based design, embedded-systems practice and experimental validation.
OutputEarly research-project foundation and competition recognition.
Level of detail: This page is designed as a quick portfolio selection. A more technical brief, demo flow or collaboration note can be prepared for a specific context.