Research
AI Security & Privacy
Robustness against adversarial attacks, prompt injection defence, learning without sharing data, and model confidentiality.
What we work on
- 01Robustness to adversarial examples and measuring performance under attack
- 02Prompt injection and data exfiltration defences in language models
- 03Federated learning and training models without centralising data
- 04Differential privacy and preventing training data from being recovered from the model
- 05Protection against model theft and reverse engineering
Why we work in this area
Putting a model into production means putting its attack surface into production too. Vision models can be fooled by perturbations invisible to the eye; language models can be steered by the text handed to them. These are not theoretical; they are things we meet in the field.
The second issue is learning without the data ever leaving. Organisations that cannot develop models for data security reasons often realise late that the problem is not the model but the architecture.
From our research projects
Our threat detection and data protection project intersects directly with this area: a defence approach that does not rely on known signatures, works from behaviour, and adapts itself to new forms of attack.
Sectors where we apply it
Related posts
The new Machinery Regulation and machine learning: for manufacturers using AI in safety functions
The rules that the EU and Turkish machinery regulations applying from 20 January 2027 set for machine learning: which components are in scope, conformity assessment, logging and boundary requirements, and model updates.
The current EU AI Act timeline: what Turkish manufacturers need to be ready for, and when
The dates changed by the Omnibus regulation, how the Act reaches companies outside the EU, which industrial systems count as high-risk, and what that means on the engineering side.
Other research areas
Generative AI & LLMs
RAG architectures, domain adaptation and fine-tuning, distillation into smaller models, Turkish evaluation sets and on-premise deployment.
Computer Vision
Object detection and tracking, video understanding, vision-language models, document intelligence and OCR, real-time inference on edge devices.
AI Agents
Tool use, multi-step planning and multi-agent orchestration; MCP integration, human-in-the-loop workflows and safety guardrails.
MLOps & Model Operations
Evaluation infrastructure, observability, hallucination and regression tracking, continuous retraining.
Reinforcement Learning & Decision Systems
Decision policies that adapt to changing conditions, reward design, training in simulation and safe transfer to the field.
A technical assessment for your AI project
Your project's feasibility, risks and timeline are assessed in a technical consultation.