AI & Software
Agents, models, inference economics, vertical AI, AI security. The layer where compute's value actually gets captured.
Agents
AI agents, orchestration, agent identity and KYA, agentic payments, x402, stablecoin rails. Software that acts — and the trust and money plumbing it needs.
Models & inference
LLMs, open-weight models, post-training, RL, inference economics, KV cache, state-space models, the CUDA moat. What the models are and what they cost to run.
AI applications
AI for EDA, diagnostics, materials discovery, simulation, physics, geospatial. Vertical AI where the domain is the moat.
AI security
Workload sandboxing, hypervisor isolation, microkernels, unikernels. Running untrusted intelligence safely.