I am an independent researcher and third-year cybersecurity student working across machine learning, systems, and distributed architectures.

My work focuses on how intelligent systems behave under adversarial pressure, long-term adaptation, and real-world deployment constraints. I study attacker–defender dynamics in learning systems while also exploring the systems-level realities that shape them — runtime environments, distributed coordination, and failure modes.

My recent work introduces Reversible Neural Adaptation (RLAE), a structural paradigm that separates behavioral learning from model identity, enabling deterministic rollback and addressing irreversible behavioral drift in weight-based systems. This work formalizes concepts such as structural irreversibility, recoverability, and behavioral divergence.

Alongside this, I design and build systems that host and stress these ideas in practice — from kernel-level thinking and runtime safety to distributed multi-agent infrastructures. My work spans AI security, adversarial machine learning, and the engineering of controllable, observable, and fault-tolerant systems.

I am particularly interested in the intersection of:

• learning dynamics under adversarial conditions
• systems and runtime constraints shaping intelligence
• distributed autonomous agents and coordination
• safety, control, and recoverability in adaptive systems

My goal is to contribute to the design of systems that are not only intelligent, but structurally reliable, observable, and governable — from model behavior to system runtime.

• Building AADS — Agentic AI Defense Swarms with safe governance and swarm-level autonomy
• Engineering runtime safety systems in Rust — kill-switches, isolation layers, fault boundaries
• Designing distributed agent runtimes in Go/K8s — CRDs, orchestration, gossip protocols
• Adversarial stress-testing MARL agents and behavioral LoRA modules
• Prototyping GNIM — cyber-geospatial intelligence mapping system
• Researching RLAE / reversible learning systems
• Papers:
  • On the Structural Limitations of Weight-Based Neural Adaptation... — arXiv
  • Formal Theory of Reversible Behavioral Learning — in progress

• Machine Learning (adaptive & robust systems)
• Systems & Distributed Architectures
• AI Security & Adversarial ML
• Cyber-Physical & Autonomous Systems

Aiming toward MSc → PhD focused on ML Research & Security, Systems & Adaptive Intelligence.

If my research or experiments contribute to your projects or spark ideas, you can support my work here:

"We are merely vessels of a greater curiosity"