AppSec MCP server
Reference implementation — 8 tools, 5 ordered guardrails, 3 pluggable LLM providers, end-to-end OpenTelemetry observability.
Portfolio
Published platform engineering and application security work — the reference implementations and architecture write-ups behind the GenAI/AppSec advisory practice.
GitOps on AWS EKS
Enterprise GitOps platform on EKS — multi-cluster DEV/PROD isolation, ArgoCD, and security controls.
GitOps platform architecture
GitOps platform architecture suitable for high-volume financial services environments.
Backstage developer portal
Backstage internal developer portal — self-service platform for a regulated enterprise.
Platform engineering evolution
From reactive operations to self-service — Crossplane, Temporal, and AI-augmented developer experience.
QuestDB on Azure
Tiered-storage architecture diagrams — QuestDB Enterprise feeding Azure Blob with declarative lifecycle management.
GitHub → GitLab migration
Migration guide and approach for moving from GitHub to GitLab.
Architecture review
Architecture review of the GitOps demo environment.
Enterprise observability
OpenTelemetry metrics, traces, and logs unified in Grafana with Tempo and Loki backends. A parent span per request, child spans for each guardrail and downstream call, and a shared trace ID injected into every log line — one correlation ID ties spans, traces, and logs together.
Claude Skills & context engineering
Field notes on progressive disclosure — how Claude Skills change the economics of agentic coding work.
Claude Code implementation field notes
Part two — wiring the playbook across a polyglot platform: memory load order, path-scoped rules, conditional guards, and managing the live context budget.
Amazon Q Developer POC
Structured 4-week proof of concept — org setup, repo-trained customizations, baseline metrics, and a go/no-go decision package.
AI coverage — tests only, production untouched
91% SonarQube coverage using Claude Opus 4 + GitHub Copilot with three enforcement layers keeping production source frozen. Reference config kit included.
Copilot coverage workflow
A Copilot-only, cost-aware workflow that raises Java test coverage to 91% — autonomous agent loop, defense-in-depth guardrails, and token engineering.
Token cost optimization
Field notes on the economics of agentic AI: why token cost grows every turn, and the four levers that cut it — lean context, load-on-demand, compress inputs, keep tool output off the meter.
Lean ML/Ops starter
A migration, guardrail, and AWS-baseline starter for a team standing up a new ML/Ops environment — built so enforcement doesn't depend on which AI tool an engineer picks.
Context & Token Cheat-Sheet
GitHub Copilot vs Claude Code — how each tool builds context, what it ignores vs loads, and the levers that control token usage.
The Credit Fire — Copilot spend parable
A Phoenix Project-style narrative on why token cost is a work-in-progress problem, not a model problem. Six levers for controlling Copilot spend, in order.
Two meters for your AI coding tools
What your AI coding assistant costs vs whether the services are up — in plain terms, applied to Copilot and Claude.
GitLab Orbit & Agentic AI
What Orbit's SDLC knowledge graph does for AI agents like Claude, where it helps, and the blind spot — the code it won't read — to plan around before you rely on it.
ML/AI Capability Roadmap
From a single ad-hoc workload to a governed, enterprise-wide platform — five maturity levels, the enterprise value chain, and responsible AI built in from Level 3.
ML/MLOps Platform Advisory — Reference Architecture
From legacy delivery to a team-owned, self-service platform — agentic-native and compliant by default. Layered architecture with phased migration path.
Technical leadership profile
Executive-director-level background spanning platform engineering and security.
