One team, from research to application
Piksel Bayt Teknoloji Limited Şirketi is an AI R&D company operating at the Istanbul Medeniyet University Technopark. Our work runs along two tracks: on one side we carry out publicly funded research projects and develop our own models and methods; on the other we turn that accumulated knowledge into products that work for organisations.
These two activities are not divided across separate teams. The team conducting the feasibility assessment is also responsible for developing the model and monitoring it in production. This structure allows the field validity of research findings to be verified in the shortest possible time.
Outside our scope
Our activities are not limited to advisory services or report delivery, and integration work based on connecting off-the-shelf services falls outside our scope. The solvability of a problem is first assessed through our own research, after which the solution is developed end to end.
R&D approach
The projects conducted are publicly funded research projects that have passed independent peer review. This indicates that the originality of the problems addressed and the scientific adequacy of the applied method are subject to external audit. The same discipline is applied to commercial projects: an evaluation set is defined for every model and a measurable metric for every performance claim.
Support and programmes
Our projects are carried out under TÜBİTAK and KOSGEB support programmes. The company holds the Teknogirişim Badge and operates within Medeniyet Teknopark. We are members of the NVIDIA Inception and Microsoft for Startups Founders Hub programmes.
Location
Medeniyet Teknopark Üsküdar AlanıÜnalan Mahallesi, Ünalan Caddesi No:1, A Blok D:2-DÜsküdar / İstanbulResearch
- Generative AI & LLMsRAG architectures, domain adaptation and fine-tuning, distillation into smaller models, Turkish evaluation sets and on-premise deployment.
- Computer VisionObject detection and tracking, video understanding, vision-language models, document intelligence and OCR, real-time inference on edge devices.
- AI AgentsTool use, multi-step planning and multi-agent orchestration; MCP integration, human-in-the-loop workflows and safety guardrails.
- MLOps & Model OperationsEvaluation infrastructure, observability, hallucination and regression tracking, continuous retraining.
- Reinforcement Learning & Decision SystemsDecision policies that adapt to changing conditions, reward design, training in simulation and safe transfer to the field.
- AI Security & PrivacyRobustness against adversarial attacks, prompt injection defence, learning without sharing data, and model confidentiality.
A technical assessment for your AI project
Your project's feasibility, risks and timeline are assessed in a technical consultation.