la backdoor is a security research lab working on data-driven solutions for threat detection and resilience in cyber-physical systems. We combine machine learning, NLP/LLMs, and classical security analysis to build defenses that hold up against real adversaries.
Research areas
- Threat detection & CPS resilience — data-driven detection and resilience for cyber-physical and networked systems.
- AI security — applying (and stress-testing) machine-learning and large language models for security, including byte-level models for network analysis.
- Network & protocol intelligence — deep packet inspection and protocol-level analysis driven by learned representations.
- Security analytics — forecasting software-vulnerability trends and mining security signal from large, noisy data sources.
Work with us
Our code and datasets live on GitHub at github.com/LaBackDoor. Contributions and collaborations are welcome — open an issue or a pull request, or reach out via the contact page.