
Edge Applications
See allAgentic AI
See allAreas of Expertise
- Agentic AI — Agentic AI, AI-native prodcuts, MCP servers, LLM orchestration on AWS Bedrock — including the CEO Innovation Award-winning 3-agent platform at Belden
- Industrial Edge & OT/IT Convergence — IIoT architecture, Margo open standard, edge deployments, MQTT, and industrial automation (SPS 2025, Linux Foundation)
- AWS Cloud Architecture — Serverless, CDK, cost optimisation, and Well-Architected patterns — backed by all 10 AWS certifications and AWS Community Builder status
- Data Lakehouse & Data Mesh — Apache Iceberg, DuckDB, Data Infrastrcuture for AI-Native Products
- Generative Engine Optimisation (GEO) — Static site generation, AI-accessible content infrastructure, llms.txt, MCP server exposure, and structured data for LLM discoverability
Latest Research
Why My Blog's MCP Server Has No Authentication
Exploring the architectural decision to skip authentication for a public blog MCP server, prioritizing frictionless access for AI agents and why public content deserves a public API.
Why Quadlet Is Different, Part 3: Production Lifecycle on the Factory Floor
Push-based readiness, kernel watchdog supervision, two-layer shutdown timeouts, hardware-coupled dependencies, and more — lifecycle control Compose can't express
Why Quadlet Is Different, Part 2: Defence-in-Depth Cybersecurity at the Service Level
Quadlet exposes systemd hardening — filesystem isolation, kernel surface reduction, network microsegmentation, syscall filtering — declared inline, scored numerically, mapped to IEC 62443
Why Quadlet Is Different, Part 1: Resource Governance That Compose Can't Express
Podman Quadlet inherits 100+ systemd directives for memory throttling, CPU pinning, I/Om and hierarchical resource budgets — capabilities Compose can't express
From AI Intern to AI Employee
Migrate your AI Agent from native Bedrock Agents to Strands Agents SDK with AWS AgentCore. Learn why ECR containers beat Lambda for agentic AI workloads










