Projects

2026

oss-model-playbook

Enterprise playbook for running open-weight LLMs on on-prem Kubernetes, AKS, and Azure AI Foundry — model selection, capacity planning, gateway, security, and TCO.

pi-bench

Composed-defense benchmark for prompt injection that grades entire defense stacks on attack success rate, false positives, latency, and cost.

Vector engineering for agents

Staff-level deep dive on embeddings, ANN, hybrid retrieval, agent memory, evaluation, and production ops for RAG and AI agents.

octerse

A token-saver bootstrap for repos that drops focused Copilot instruction files and VS Code defaults.

2025